Content uploaded by José M. Ciordia
All content in this area was uploaded by José M. Ciordia on May 06, 2022
Content may be subject to copyright.
and Their War Anchors
The Naval Origin of Greek Architecture and Sculpture
José M. Ciordia
Abstract: The many surprising features of Ancient Greek temples are understood as a
matter of course if we interpret that temples originated in real warships, not metaphoric
ones, overturned and stored upon supports; there are enough ethnographic parallels for
such a practice. As well as certain references in literary sources and in vocabulary, the
definitive argument that supports this claim is the comparison between penteconters
from the Geometric period and Doric and Ionic temples: the shape of the whole
entablature, the window aspect of metopes, the shields hanging on the architrave, the
longitudinal light curvature, the size of the temples and the distribution of their interior,
its axial symmetry, the creation of peristasis, the number and distribution of columns,
the placement of sculptures and the main worship statue, etc. The principal statue comes
from the sacred anchor, which would be used as a projectile in the naval battles in the
Aegean Sea for centuries before the invention of the ram; this practice can be traced in
many heroic legends. Greek temples are more than buildings: they are sculptures in
stone of beached ships which proclaim the Greeks’ familiarity with the sea and their
mastery of it. In other words, their thalassocracy.
Keywords: penteconter; Greek temple; peristasis; naval artillery; nautegology.
Draft: May 6, 2022.
ἐὰν µὴ ἔλπηται ἀνέλπιστον οὐκ ἐξευρήσει. ‘If you do not expect the unexpected, you will
not find it’. Heraclitus (fr. 18 DK)
1. The problem and a proposal
Various characteristics of the classical Greek temple defy logic. (a) With the exception of
statues for worship, architectural statues and so-called ‘ornaments’ are many metres high,
where they look small, deformed by a low angle perspective and, in the case of the frieze of
the cella, in very little light. (b) Peristasis doesn’t seem to have any particular function, either
utilitarian or symbolic. (c) The same applies to entablature: contrary to Vitruvius, triglyphs
cannot represent the ends of the cross beams, because these do not terminate in those, except
further up, behind the cornice (figure 31). (d) In the more sophisticated temples, such as the
Parthenon, there are very few straight lines, which has been put down to certain supposed
optical corrections; these supposed corrections are so unlikely that no subsequent architect has
needed to use them again. (e) The temple arises with no relation to domestic architecture, and
with no human function other than to place and exhibit a sculpture and other objects of value.
(f) Moreover, the temple arises with a very standardised model, the hekatompedon temple,
without a lengthy phase of prior experimentation with disparate local traditions, as would be
expected. (g) Lastly, even though most of Greece lies near the sea and classical temples
appear to have a particularly close relationship with it, we do not really know what that
All the aforementioned difficulties can be resolved by forming a hypothesis that at
first appears rather daring: that the Greek temple may be the representation in stone of an oar-
propelled warship, a penteconter, flipped over and placed on walls constructed to serve as a
support, as represented in figure 1.
Figure 1. Hypothetical penteconter agrounded. Drawing by the author. License Creative Commons BY-
NC-SA 4.0 International.
Although this hypothesis benefits from the advances of the last fifty years in our
knowledge of ancient naval construction, it is not a consequence of any specific
archaeological find, but rather a new interpretation from well known evidence present in
written sources, vocabulary, archaeology and the history of art, some dating back many years,
other even centuries.
The similarity between ναός ‘temple’ and ναῦς ‘ship’ has always attracted the attention
of students of Greek. And also that of scholars: Indra McEwen, describing the Heraion of
Samos, writes that “the shrine of the goddess became a ship” and that it had—referring to the
peripteral colonnade—“well-fitted oars (…) that are as wings onto ships”. However, this
The general lines of this thesis were published by me in draft form on the internet (Ciordia 2005a and
2005b), as can be seen at https://web.archive.org/web/2005*/pompilos.org. Years later, Pantazis
(2013) published the same thesis, providing almost the same arguments without citing my
McEwen 1993, p. 101. The passage of McEwen includes a citation of Od. 11.120.
author interprets the similarity between temple and ship in metaphorical terms, rather than the
strictly literal ones that I will suggest here. On the other hand, John Onians explains how a
temple’s central row of columns and the peristasis of Ionic temples are similar to a warship,
but attributes the similarity explicitly to a “subliminal association [of the temple] with a
But before comparing an ancient Greek temple and a ship, it is necessary to rule out
three reasons for incredulity: if all of this is true, why don’t the ancient sources say anything
about it? How would someone overturn and lift a vessel that weighed various tonnes? And has
anyone ever used an overturned vessel as a roof?
The first two are lesser difficulties and I will dedicate little space to them. In the
written sources, much information is missing because what is obvious to the author and the
recipient is not made explicit. Furthermore, in science it is usual to find what we are looking
for, and as far as I can see, nobody has read the classical texts to find evidence there that a
temple is an overturned ship. In any case, even though this article will focus on the physical
comparison of temple and ship, I will also provide some references from the written texts and
It is true that it would not be easy to overturn a craft like a penteconter, which stripped
of all its equipment would weigh around five tonnes. But it could be done. Penteconters, like
vessels of every age, often sailed in flotilla. This means that in order to upturn and lift the
vessel, we would not be talking about the strength of fifty crew-members, but instead double,
triple, or quadruple this figure. In addition, machinery would have been used: the
penteconter’s mast was detachable and it is reasonable to imagine that—stuck in the sand,
Onians 2002, 54.
Coates 1999, 111.
tilting the yard (the horizontal spar from which the sails are hung) and using ropes, pulleys
and tackle habitual in a ship—it could be converted into a crane which, combined with two,
three or four other cranes from the same flotilla, would facilitate the manoeuvre.
2. Ethnographic parallels
But, has anybody ever used an overturned vessel as a roof? On beaches all over the world
there are rowboats beached in the open air. Their owners drag them onto the sand, at a
distance from the water, and, to prevent rainwater from filling them and rotting their wood,
turn them over. And that is because an overturned boat’s hull forms a natural roof: rainwater
falls down both sides of the keel, much like the ridge of a roof, and is then drained by the
edges of the hull, which act like a cornice. The craft’s impermeability prevents the rainwater
from leaking through the planks and dripping inside. An overturned hull fulfils this function
so well because that is what it does in the water—prevent sea or river water from entering the
vessel’s interior. Indeed, this is the reason why even today in the West architecture and naval
architecture share name, training, and professionals.
A boat overturned this way can accumulate dampness in its interior, and its wood can
end up rotting. Therefore, it is normal that its owner raises one or both sides of the hull and
rests them on large or medium-sized stones. Thanks to this opening the air circulates inside
and keeps the interior dry. When a sailor on the beach with his boat needs shelter from the
wind and rain, he needs only to enter this opening and take refuge inside. In April 1916 the
members of Edward Shackleton’s failed Antarctic expedition arrived at the frozen and isolated
Although writing some time later, Aeschylus testifies to the use of winches for hauling the ships
ashore (Suppliants 441); in the 1st century BC Vitruvius describes some machines, such as
winch, crane and polyspaston (10.2.1–9), and specifies that they were used for hauling ships
Elephant Island, in two boats. Two sailors suggested overturning the boats, setting them on
small walls five spans high and using tarps on the sides (figure 2). They camped this way for
four months until they were rescued in a safe and sound condition.
Figure 2. The marooned twenty-two men's camp on Elephant Island. Photo: Frank Hurley 1916.
https://www.ngv.vic.gov.au/explore/collection/work/8554/ Courtesy of the National Gallery of Victoria,
Eskimos tended to protect themselves from the summer winds by camping in the
safety of their umiaks (figure 3), the spacious ‘women’s boats’ used for transporting cargo:
they overturned them and rested one side on the ground and the other on the oars or on
Alexander 1998, p. 170.
Fair 2005, p. 233.
Figure 3. Eskimo family sitting beneath overturned umiak, Alaska, c.1904. Photo: Frank H. Nowell.
Collection of the University of Washington (NA2153). Public domain.
On other occasions a vessel became a building only at the end of its life, as a form of
re-use. This is the case with the three herring boats that the architect Edwin Luytens turned
into sheds on the island of Lindisfarne (England) in 1908, creating a local tradition that
persists to this day (figure 4). In the first half of the 20th century, a fishing boat converted into
a dwelling could be seen on the island of Lambay, in the Irish Sea (figure 4b), and at the turn
of the 19th and 20th centuries in Équihen-Plage, on the French shore of the Strait of Dover,
there was a whole district called Quartier des quilles en l'air inhabited by fishing families
who had built their houses out of discarded boat hulls (figure 4c). Also in literature, Charles
Dickens had one of the characters in David Copperfield, the fisherman Mr Peggotty, live in a
beached ship converted into a dwelling, which the illustrator H.K. Browne depicted as
Dickens 1850, p. 62; illustration by Browne to face page 186.
Figure 4. Upturned boats in Lindisfarne harbour in use as sheds. Photo: Russ Hamer. <https://
commons.wikimedia.org/wiki/File:Lindisfarne_Upturned_Boats_in_Harbour.JPG>. License Creative
Commons BY-SA 3.0.
Figure 4 b. House made of old boat on Lambay Island. Photo: Maurice Curtin (ca. 1935). <https://
doi.org/10.7925/drs1.duchas_548> Dúchas © National Folklore Collection, UCD is licensed under CC
Figure 4 c. FishermanFamily (Keels to sun district, Équihen-Plage). Photo: Anonymous. <https://
KGacY9cvb7DqG6_F.jpg> Public domain.
In some vernacular styles of architecture there are buildings that, despite having roofs
in the form of a ship, obviously never sailed. These structures are of maritime origin, either
because they were inspired by real ships that at one time served as dwellings, or because the
forms and procedures developed in naval construction were transferred to architecture on
land. In the North Sea, from the 8th to the 11th century, the Vikings developed a prolific
architecture of boat-shaped long houses of which have been preserved not only the floor plan
but also enough evidence to enable us to conclude that they had roofs shaped like overturned
hulls (figure 5). Much further afar and centuries later, among the natives of Easter Island, the
most common type of house before the arrival of the Spanish was the hare paenga, a house
for an extended family shaped like a great canoe, lengthened and turned over (figure 6). For
their part, the Torajans of the island of Sulawesi (Indonesia) still build their traditional houses,
the tongkonan, with an oversized roof in the shape of a boat. According to legend, they
reached this island in ships which, when they decided to settle permanently, they used as their
Figure 5. Reconstructed Viking Farm at Bukkøy (Norway). Photo: Frode Bjørshol. <https://
commons.wikimedia.org/wiki/File:Vikinggarden_L1020599.jpg>. License Creative Commons BY 2.0.
Figure 6. Reconstructed hare paenga in Eastern Island. Photo: והיבא (cropped). <https://
commons.wikimedia.org/wiki/File:Akahanga_Longhouse1.jpg>. License Creative Commons BY-SA 4.0
Julistiono and Arifin 2005, p. 2667.
Architects today, though surely unaware of the tradition explained here of ships that
end up converted into buildings, design structures or roofs in the shape of an overturned
vessel to fulfil some expressive need as is typical of architecture understood as art. Finnish
architect Matti Sanaksenaho conceived St Henry’s Ecumenical Art Chapel, built in Turku in
2005, in the shape of both a fish and an overturned hull, inspired by the fact that the fish is a
primitive symbol of Christ and that a boat expresses ‘the idea of community travelling
towards the same goal’ (figure 7). We will see later how the three processes that we have
documented in the preceding paragraphs—camp, re-use, and imitation, respectively—
intertwined in the creation of the classical Greek temple.
Lavalou 1998, p. ??
Figure 7. St. Henry's Ecumenical Art Chapel by Matti Sanaksenaho. Photo: Htm. https://
commons.wikimedia.org/wiki/File:Ecumenical_Art_Chapel_of_St._Henry_in.jpg Licencia Creative
Commons BY-SA 3.0.
The Eskimos, English, Rapa Nui and Vikings shared with the ancient and present-day
Greeks their dedication to maritime navigation and a close familiarity with boats. And thanks
to their example we have converted into likelihood the idea—initially outlandish—that the
ancient Greeks were able to turn over their ships and take refuge under them. But it is
necessary to demonstrate that they really did do it.
3. Clues in the ancient Mediterranean
During the Bronze Age there arose, in the Balearics, a culture known as the Talaiotic. One of
the most characteristic buildings of this culture is the naveta (figures 8 and 9), ‘small boat’ in
Catalan, which was used between the twelfth and tenth centuries BC, often as a place of
habitation, other times as places of collective burial. These stone structures were named so
because their shape recalled overturned boats.
In the Eastern Mediterranean, and the end of the Bronze Age, the so-called ‘Sea
People’, a mix of mariners from the Western Mediterranean and the Aegean (Sardinians,
Sicels, and Mycenaeans, among others), attacked the Hittite Empire, the Levant and Egypt.
Referring to one of these peoples, most likely the Sicels, the king of Hatti, in a letter to the
king of Ugarit, used the eloquent expression ‘šikalayu, who live on ships’. A similar
expression is used in the Song of Deborah (Judges 5:17) in reference to the tribe of Dan,
which according to some scholars derived from the denyen, another of the Sea Peoples:
‘And Dan, why does he live in ships?’.
Centuries later, in Lycia, the coastal region situated in the southeast of Asia Minor and
firmly Hellenized, notables were buried in what until now have been considered stone
sarcophagi, some free standing (figure 10), and others sculpted into the cave entrance. In my
opinion these structures do not represent sarcophagi, but instead the hulls of overturned ships,
as would befit the sea captains of a nation dedicated to maritime navigation.
Furthermore, during the first millennium the Egyptians used on the Nile and the
Eastern Mediterranean a type of shallow draught vessel that Greek sources call βᾶρις.
Guerrero, Calvo and Gornés 2006, 107–113.
Dietrich and Loretz 1978, 53–56.
Yadin 1968, 17–23.
Singer (1988, 246) translates it as ‘why does he live on ships’, although usually the expression is
translated as ‘why did he remain in ships?’.
In the classical period, Aeschylus calls a ship “wooden house sewn with linen that shelters from the
sea” (Suppliants 134–135) but, as the author is a poet, maybe we are merely dealing with a
Curiously, the same name designates, in the Septuaginta (Psalmi 44.9, Daniel 8.2), a
communal house that could accommodate a company of soldiers.
Figure 8. Naveta d’Es Tudons, Menorca (Spain). Photo: Isolmar. <https://commons.wikimedia.org/wiki/
File:Naveta_des_Tudons.jpg> License Creative Commons BY-SA 3.0 Spain.
In the same line, Stephanus Byzantinus informs (s.v. Βάρις): Λέγεται βᾶρις ἡ οἰκία, ὡς Ποσείδιππος,
καὶ ἡ συνοικία ὡς Ἔφορος. ‘Baris means house, according to Posidippus, and communal house
according to Ephorus’.
Figure 9. Navetes at Cala Morell, Menorca (Spain). Photo: Equip Entre Illes. <https://
commons.wikimedia.org/wiki/File:Es_Coll_de_Cala_Morell.jpg>. License Creative Commons BY-SA
Figure 10. The tomb of Payava, from ancient Lycia, present-day Turkey (British Museum). Photo:
Carole Raddato. <https://www.ﬂickr.com/photos/carolemage/9502129675/>. License Creative
Commons BY-SA 2.0.
The four previous examples suggest that overturning vessels in order to take shelter
underneath could also have been a common practice in the ancient Mediterranean. The most
literal form of this (a wooden boat resting on wooden supports) does not appear in the
archaeological record as such. Apparently, the Mycenaeans’ contact with the inhabitants of the
Balearics was more indirect than direct. They did have contact with the Sea People, to
whom the Mycenaeans may be added, and obviously with the Lycians, whose region was a
maritime transit point in the voyage to and from the Levant and Egypt.
The Mycenaeans and, after the lapse of the Geometric period, the Greeks of the
Archaic period sailed around the coasts of the Mediterranean in cargo vessels, as well as
penteconters. There is abundant evidence that these light oared warships were often beached
by hauling them out of the water, especially when they were going to be unused for a long
time. Seafarers settled in some of the regions they arrived in for periods of varying duration,
that would sometimes include spending all winter in contact with the native population, but
without mixing with them. If they camped beneath their penteconters, the appearance of such
constructions would be that shown in figure 11: a compartment some 38m long and 4m
wide, with the entrance at one end, the stern, orientated preferably to the east for reasons of
climatic comfort, with at least one curved end, the bow, and a large stone fireplace situated
outside but close to the door in order to avoid setting fire to the wood. Lacking benches and
tables the sailors would eat and sleep lying on their rowing benches. Regarding both
constructions, the bunkhouse formed by the overturned ship, and the stone fireplace for
Guerrero, Calvo and Gornés 2006, 238.
See a recent review in Votruba 2017, 7–14.
The rowers of the monokrotos penteconter, as its name indicates, were all on one level. It measured
125 feet and length and 13 in width. The dikrotos ‘of two levels’ was half that in length (65 feet)
but was similar in width. Casson 1995, 54-56.
cooking, the first would hardly leave a trace of its post holes in the soil. The latter, however, is
more durable. This is probably the reason why in various archaeological sites the stone altar
precedes by decades or even centuries the first real buildings: in the sanctuary of Hera at
Samos the altar dates from the end of the Mycenaean period, and the first temple from the
8th century BC. It is worth noting that the term ἐσχάρα ‘fireplace, altar’ has a derivative
ἐσχαρέυς ‘altar attendant’ that is used twice in the written sources (Poll. Onom. 1.95 and
Them. Or. 15.195b), in both cases applying to ships’ cooks.
Figure 11. Hypothetical penteconter used as a shelter and outdoor ﬁreplace next to the entrance.
Drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
The Iliad is not a work of history, but rather a source of ethnographic information that
shines a light on how the Greeks lived in the 7th century BC and earlier. In the Iliad, the
Greeks live for years in a camp before the city of Troy, a camp which includes the ships and
the barracks in which they took shelter. It is difficult to know exactly what the arrangement of
the ships and barracks was, but they must have been very close to each other because Homer
frequently describes them as forming a unit: Μυρµιδόνων δ’ ἐπί τε κλισίας καὶ νῆας ἱκέσθην, /
τὸν δ’ εὗρον παρά τε κλισίῃ καὶ νηῒ µελαίνῃ / ἥµενον· (‘…they came to the barracks and ships
of the Myrmidons. / They found Achilles sitting by a his barrack and his ship.’) (Il.
Kyrieleis 1993, 128.
1.328-330). The word ship is sometimes used in place of barracks: λῦσεν δ’ ἀγορὴν
αἰψηρήν. / οἱ µὲν ἄρ’ ἐσκίδναντο ἑὴν ἐπὶ νῆα ἕκαστος (‘He broke up the assembly, and every
man went back to his own ship.’) (Il. 19.277). Finally, on one very revealing occasion, ship
and barrack are used as equivalents. Agamemnon decides to make up for offending Achilles
by giving him gifts, and asks Odysseus (Il. 19.190-194):
…µίµνετε δ’ ἄλλοι πάντες ἀολλέες, ὄφρά κε δῶρα
ἐκ κλισίης ἔλθῃσι καὶ ὅρκια πιστὰ τάµωµεν.
σοὶ δ’ αὐτῷ τόδ’ ἐγὼν ἐπιτέλλοµαι ἠδὲ κελεύω·
κρινάµενος κούρητας ἀριστῆας Παναχαιῶν
δῶρα ἐµῆς παρὰ νηὸς ἐνεικέµεν…
…do you others wait also, till the gifts
come from my barrack and we ratify the oath with sacrifice.
Thus, then, do I charge you:
take some noble young Achaeans with you,
and bring from my ship the gifts…
There is no reason, apart from our own conceitedness, to think that Homer didn’t
know what he was writing, or that he was making a mistake. We can conclude that the poet
used both words interchangeably because both objects formed part—the barrack below and
the overturned ship above—of the same temporary architecture. In any case, there should be
noted that Homer clearly states in Il. 24.448–451 that a hut specifically made for Achilles, not
a comunal one, is crowned with reed roof.
The names used to refer to the classical temple presents a particularity that may be
related to the configuration we have proposed. There is no specific term for it: the too generic
to hieron ‘the sacred’ is sometimes used, and elsewhere the too specific ho naos, which
strictly speaking designates the enclosed part of the classical temple, the cella in Latin,
All quotes from the Iliad and the Odyssey are taken from Butler (1898 and 1900 respectively), with
some lexical correction.
excluding both the temple’s ceiling and the peristasis. This lexical peculiarity can be
explained as follows.
The enclosed part of the temple is sometimes called ὁ ναός but also ὁ οἶκος, which has
the generic meaning of ‘the building’. It is likely that both derive from a single syntagma ὁ
ναὸς οἶκος, which means ‘the ship’s house’ or ‘the naval house’. The former translation
assumes that ναός is the genitive of the noun ‘ship’, while the latter assumes that ναός is the
nominative singular of an ancient adjective derived from ναῦς, meaning ‘naval’, which later
became a masculine noun meaning ‘temple’; in Greek both values of ναός are
indistinguishable except for the context. In Classical Greek, the syntagma ὁ ναὸς οἶκος gives
rise to the compound noun *ναόσοικος > Attic νεώσοικος, one of the usual terms for
‘shipshed’, the structure where ships were stored and repaired. As a matter of fact, the
similarity between naus and naos that caught the attention of students of Greek is no
coincidence. Rather, it is proof of an etymological relationship: ὁ ναός originally meant ‘the
naval’, a construction used to support a ναῦς ‘ship’, or, strictly speaking, a roof.
Sailors used overturned ships to make camp while overseas, such as on the southern
coasts of Italy and Sicily before establishing colonies there. However, back in Greece, each
sailor would return to the family home. The ships, stored outside urban centers and overturned
in order to better survive the winter, would then be put to various other uses: some would be
employed as storehouses for the ship’s equipment, others as winter stables for livestock, and
Scholars have maintained that there is no etymological relation between the terms, and that ναός
derives from the proto-Greek form *naswo-, which has no relatives in any other language, Indo-
European or otherwise (Beekes 2010, s.v. ναός). The Mycenaean na-wi-jo of the tablet Jn 829
from Pylos can be a derivative of naus or naos (Del Freo 2005, notes 28 and 29).
The word ἡ σηκός ‘pen, fold’ also means ‘sacred enclosure’ of the temple and ‘bedroom’. The three
meanings are coherent with the hypothesis of the naval origin of the Greek temple presented in
a few others as meeting places and venues for occasional banquets by the captain and crew.
These uses probably gave rise to stoas, temples, and other posterior constructions which were
so similar in form.
But written sources and archaeology also testify to the dedication of ships to the gods
in sanctuaries. Herodotus (8.121) claims that after the Battle of Salamis three Phoenician
triremes were offered up, one in the sanctuary at Isthmia, another in Sounion and another in
Salamis, and that the first of these could still be seen in his time. In the sanctuary of Hera at
Samos, we can still see the stereobates that supported an Archaic period ship of 23m in length
and between 3 and 4m in breadth. In the sanctuary at Delos we find the striking Neorion
(Arsenal), a very long building from the 3rd century BC that housed the flagship of Antigonus
II Gonatas, dedicated after a naval victory; the Neorion’s floor recalls those of the first
temples of the Geometric period. In the Sanctuary of the Great Gods of Samothrace another
Neorion from the same century housed a warship in its interior. Herodotus says nothing
however about how the Isthmia trireme that he saw with his own eyes happened to be stored;
if it was in its usual position or overturned. Nor does he mention its similarity to a stone
temple. For the sake of the argument here, however, it is enough to note the physical
proximity and the functional oneness—as an offering to the gods in the same sanctuaries—of
ships and temples.
Scholars claim that the Greeks copied the custom of erecting stone temples from the
Egyptians. This is surely the case, but we should also add an intrinsic motivation, inspired by
the poems of Homer, or reflected in them. In the Odyssey, Poseidon threatens to sink the ship
in which Phaeacians took Odysseus home. Zeus suggests him an alternative revenge:‘turn [the
Wallinga 1995, 37 y 40.
ship] into a rock near the land / looking like a ship. This will astonish / everybody’ (Od.
13.156-158), action that Poseidon finally performs. According to this passage, the dedication
of a stone ship, maybe a temple, would be a deal with divinity to ensure the ships’ safety and
the continuity of the maritime endeavour.
All the aforementioned suggests that Greek temples may have originated in stored
ships, but the definitive evidence for this rests in the formal similarity between a classical
temple and a penteconter placed on stanchions. A sleeping penteconter, as it were. I will begin
the comparison with the entablature, which must surely be, since Vitruvius, the least
understood part of the Greek temple.
4. Doric entablature as a gunwale
Archaic epoch ceramics were often decorated with drawings of ships. Maritime
commerce was the most lucrative economic activity of the age, and that which gave rise to
Greece’s resurgence from the 8th century BC.. Thanks to these images we know the form of
the penteconter, a ship that was propelled by fifty oarsmen and typical of the time even
though archaeologists still have not found or excavated a single wreck of this type of vessel.
In the ostrakon in figure 13 (LH IIIC, circa 1200-1090 BC) we see firstly that the oarsmen
were arranged on a single level. Secondly, that above them was a screen that ran from bow to
stern, which Homer calls αἱ ἐπηγκενίδες ‘gunwale’ (Od. 5.253). And finally that between the
hull and that screen were a gallery of windows, one per rower, through which the oar
Murray, trans., 1914, with a small correction.
There were also dedications of stone ships clearly recognisable as such. According to Procopius
(Goth. 4.22.23-29), in the 6th century, there was a ship of white stone at Cassopa, on the island
of Corcyra, which some believe has a relation with a Phaeacian ship. There was another in the
port of Geraestus in Euboea which was dedicated to Artemis Bolosia.
protruded and the rower could be seen. The bas-relief in the palace of Senacherib that
depicts the evacuation of King Luli of Tyre in 701 BC (figure 14) shows Phoenician ships
with two levels of oarsmen from the same period in which can be seen more clearly the same
structure. This type of gunwale could be in use on Mediterranean warships throughout the
first millennium BC, if Basch is right in dating to the 1st century BC or early 1st century AD
the clay model of a galley taken out of the water in the harbour of Gytheion, in the
Peloponnese, and lacking archaeological context (figure 14b); Höckmann, on the contrary,
believes that it represents a ship from the Archaic period. On the upper screen of this clay
model there is a row of small holes in which, in all probability, shields like the ones depicted
in the Phoenician model and hung in the Parthenon were attached. Well then, if we turn these
three ship models upside down, the upper structure turns out to be identical to the entablature
of a Doric temple. The upper elongated screen turns into the architrave, the windows of each
rower’s post becomes the metopes, and the stanchions that separate the windows and hold up
the upper screen become the triglyphs.
A detailed explanation in Wachsmann 1995, 23–33. The same configuration has the gunwale of the
ships of the Sea Peoples on the relief of the naval battle of Temple in Medinet Habu (Egypt) from
1186 BC, as Wachsmann (1981, figure 12) interprets it. Years later the same author (1995, figure
p. 32 bottom) interprets the upper screen as being made of fabric or leather, but the evidence
provided by figures 16 and 17, in which the screen is a rigid pavisade capable of supporting
shields, disproves this possibility.
Basch 1987, 428 and 434-435.
Höckmann 2002, note 7.
Basch 1987, 428.
Figure 13. Ship depiction from Pyrgos Livanaton, Homeric Kynos (Greece). Drawing by © Wachsmann
Figure 14. Phoenician galleys evacuating King Luli from Tyre in 701 BC.. Palace of Sennacherib,
Nineveh. Drawing by Layard (1849, plate 71). Public domain.
Figure 14b. Clay model of a warship found at Gytheion. Archaeological Museum of Neapolis Voion
(formerly in A. M. of Sparta). Courtesy of © Rasmussen 2015.
The well-known explanation of the entablature from Vitruvius (4.2.2)—that the
triglyphs were transverse beam ends, and the architrave the longitudinal beam that supported
them—has been repeated many times for want of a better explanation. The truth is that behind
the triglyphs there is no beam: the transverse beams of the temple’s roof are located higher up,
at cornice level. As previously mentioned, the monokrotos penteconter had a length to
breadth ratio of ten to one (125 x 13 feet) and a very shallow draught. This made it extremely
long and thus very vulnerable to sagging and hogging, which could split the vessel in half.
The upper screen appears to have had two functions: to act as a longitudinal beam that made
the structure more rigid, and secondly, to protect the oarsmen’s heads. The ταινία ‘band’,
Viollet-le-Duc already pointed out this difficulty (Bowen 1950, note 6). Other objections to
Vitruvius’ explanation in Bowen 1950, 113-114.
Curving of the structure due to the changing way in which the ship sits on the waves that pass below
it, especially when the interval of the waves is equal to the length of the ship. The ship’s hull is
sagging when its centre is lower than the fore and stern, and is hogging when it is higher.
together with the architrave, would convert this into an ideal T-beam, and would also serve as
a connecting piece for the architrave and triglyphs using the bronze nails and wooden pegs
depicted in the guttae. The clay model of a warship found at Cape Maleas (figure 14b) has an
obvious molding line in the place that corresponds to the architectural tainia.
The triglyphs’ two vertical grooves appear not to have had any other function other
than to reduce the weight of the piece without compromising its rigidity. The triglyphs were
painted blue (Vitr. 4.2.2), possibly to create a continuous line when the triglyph’s blue joined
the color of sea and sky. In Euripides’ tragedy Iphigenia in Tauris (lines 111-114), one of the
characters says to Orestes:
τολµητέον τοι ξεστὸν ἐκ ναοῦ λαβεῖν
ἄγαλµα πάσας προσφέροντε µηχανάς.
ὅρα δέ γ᾽ εἴσω τριγλύφων ὅποι κενὸν
That carven image must we dare to take
Out of the shrine with all the craft we may.
Mark thou betwixt the triglyphs a void space
Whereby to climb down.
As the triglyphs mentioned here gave access to the interior of an enclosed space, we
are not talking about the triglyphs of a peristasis. We are dealing with a simple temple which
comprised only a cella, without peristasis, with the triglyphs forming part of the entablature of
The end of the grooves suggests that the piece (in reality the entire ship) is positioned in reverse: the
part below the grooves, where they meet the tainia, will accumulate dirt. However, on the ship in
its normal position, the beveled ends of the grooves would be facing downwards and would
evacuate water and dirt.
the upper part. It is likely that the expression εἴσω τριγλύφων ‘inside’ or ‘between the
triglyphs’ refers to the metopes and that these still had not been enclosed with boards.
The oar windows could be closed, probably with wooden boards (πίνακες), to protect
the oarsmen, while keeping an opening for the oars. This must be the origin of the
metopes. When the ship was overturned to convert it into the roof of a permanent building,
the surface of the pinakes would be painted to depict the oarsmen behind it as warrior
mariners—i.e., the precursors of those who built the temples, portrayed in mythology as
heroes—and their feats beyond the seas.
Another connection between temples and ships: the temples’ proportions are
calculated from minimum units that we call modules. From Vitruvius (3.2.5) we know that
one of them was intercolumniation. This practice could have come from naval architecture
where it is perfectly justified. In the design of oar-propelled ships, such as the penteconter, the
basic unit for calculating length is the interscalmium, the space between two oars that defines
the size of the ‘room’, or space occupied by each oarsman; in Antiquity the interscalmium
was that of two cubits (Vitr. 1.2.4), the Doric cubit being 0.49m, making an interscalmium of
Vitruvius (4.2.4) comments on the same matter: Non enim quemadmodum nonnulli errantes
dixerunt fenestrarum imagines esse triglyphos ita potest esse (‘It cannot be that the triglyphs
represent windows, as some have erroneously said’. Morgan, trans., 1914).
Casson claims: “the last step, it appears, was the closing in of the slot, thereby making the hull solid
right up to the level of the erstwhile raised deck” (Casson 1995, 57).
The space between the mutules is known as viae ‘steps’; the name may refer to slots for the oars but
in this location there is no moulding in the form of a slot. In the galeras in figure 14, the openings
are in the hull, below the gunwale; as in Greek ὀπή means ‘opening’, the architectural term
µετόπη can mean ‘above the opening(s)’.
Morrison 1995, 63.
Alexander the Great dedicated to Athena in the Parthenon shields of gold from the
spoils taken from the Persians after the Battle of Granicus. Shields from other battles were
hung in the entablature of the Temple of Apollo in Delphi. It was habitual for citizens to
dedicate objects of varying value to the gods in the temples. The temple itself is an ἀνάθεµα, a
dedication to the gods—but these objects were kept and exhibited inside the temples. The
placement of the shields, in both Athens and Delphi, is striking, and coincides exactly with the
placement of the oarsmen’s shields—hanging outside the gunwale—in Phoenician galleys that
represent the evacuation from Tyre of the King Luli (figure 14) as well as in many other ships
depicted in ancient imagery. That is why the sensible thing is to think that its function is
figurative, that the shields which hang from the entablature represent the shields owned by
each oarsman who rowed from the metopes and who were represented in them.
In the peristyle of some Roman domus of the 1st century AD and the first half of the
2nd, both Doric and Ionic, features known as oscilla ‘swings’ were hung from center of the
inter-columns. The origin and function of such objects has been interpreted in various ways
(Taylor 2005), but as early as 1881 Maurice Albert identified two types of oscilla as shields:
the circular oscilla would be a derivative of Roman shields of the clipeus type, similars to the
Greek aspis used by the hoplites, and others, as the one in figure 15, had clearly the shape of
Aeschines 3.116 and Pausanias 10.19.4.
In the 10th century AD an anonymous Byzantine writes about this part of the warships of his time:
Ἅνωθεν δὲ πάντων ἡ ἐπηγκενίς, τὸ ἄρτι λεγόµενον καταπατητόν: ἐκεῖσέ που καὶ τὸ καστέλλωµα
γίνεται, ἔνθα τὰς ἀσπίδας οἱ στρατιῶται κρεµῶσι (‘Above everything [is] the gunwale, known
lately as the katapateton. Somewhere here there is also the pavesade where the soldiers hang
their shields’) (Anonymous, Naval warfare, commissioned by Basil, the patrikios and
parakoimomenos 2.13; E. M. Jeffreys, trans.).
At the beginning of the 17th century the Spanish lexicographer Sebastián de Covarrubias wrote
about the etymology of the word ‘galleries’: ‘They could be called like this because of their
length and the similarity of the fenestration to the pavisade of the galley’ (1611, s.v.).
pelta (Taylor 2005, 84–84 and 93), the type which the Romans attributed to the Amazons. It
is risky hanging a stone ornament near a well-travelled route and architects would have
avoided it, except for when there was a good reason. It has been assumed that the function of
these shields was votive (Taylor 2005, 93), but in my opinion their function is figurative: on
the gunwales of penteconters the shields were hung by their straps, and in Roman peristyles
the oscilla would represent the shields that, once the boat was overturned, would hung from
their straps over the lower opening.
Figure 15. Roman Campana relief with Heracles under an Ionic peristyllum with oscilla (Ny Carlsberg
Glyptotek, Copenhagen). Photo: Daderot. <https://commons.wikimedia.org/wiki/File:Herakles_-
_Ny_Carlsberg_Glyptotek_-_Copenhagen_-_DSC09304.JPG>. License Creative Commons 0 1.0
That it was the shield of the Amazons and associated with Ionic style coincides with Vitruvius’
belief (4.1.7-8), shared by his contemporaries, that Ionic was a feminine style.
Below the gunwale, warships often had a ὑπόζωµα, a term that can be translated as
‘diaphragm’ or ‘belt’, which consisted of two thickly braided ropes that girdled the hull the
entirety of its perimeter. The function of the hypozoma was to fasten the hull, damping and
reducing the movements of expansion caused by the flexing of the whole structure. In the clay
model of a warship found at Cape Maleas (Figure 14b) the hypozoma is clearly visible, going
from the ram to the lower part of the rowing windows, where it is no longer visible. In the
temple, in the upper part of the entablature or on the lateral side of the cornice there are often
represented, among other ornaments, meanders of two or three braided lines, often in curves
but also rectilinear (figure 16). If these meanders represent a braided hypozoma, they are
giving us information that we did not know about. In each twist or braid of the ropes, what
appears to be a nail of circular cross-section, similar in form to a gutta, would keep the ropes
in position and join them to the hull, so that in case of rupture, a broken rope would stay
nailed to the hull, the break would be localised at that point and would not spread to the rest
of the hypozoma.
Figure 16. Painted meander on a terracotta geison revetment (unknown building in Akragas, 570-530
BC.). Drawing © by Rheeder (2019, ﬁg. 5.2-7).
The Greek temple has curvatures where a modern architect would use straight lines.
One of the most well-known is the very slight curvature of the entablature, which is
replicated in the stylobate and in the roof (figure 17, with exaggerated proportions). It is not
found in all temples, but only in the most refined. This curvature is associated with that which
affects the column shafts, which Hero of Alexandria (Deff. 135.14) justified on aesthetic
grounds, for correcting a supposed optical distortion which would result if straight lines were
used. But the truth is that present-day architects, when they design similar rows of columns,
do not feel the need to correct any similar optical distortion, which may well have been
To provide an example, the curvature of the stylobate in the temple of Aphaea in Aegina is up to 3
cm in 24 m, a proportion of 1/800. Bankel 1993, table 56-8.
Surely only the thickening of the angular column of the peristyle is due to an optical correction
(Rankin 1986, 38–39).
Figure 17. Schematic representation of a Doric temple with longitudinal curvature and columns
inclined towards the interior (with exaggerated proportions). Line drawing by the author based on
Coulton (1977, ﬁg. 44).
In both ancient and modern ships there are very few straight lines beyond the
centerline, which marks the ship’s axis of longitudinal symmetry, and which can only be seen
from above. The entablature’s curvature coincides exactly with the curvature known in ship
design as ‘sheer’, which affects the shape of the gunwale and, when applicable, the deck
(figure 18); this curvature is clearly visible in the clay model of a warship from Gytheion
(figure 14b). The sheer is always more pronounced in the bow and stern but both elements are
absent in the overturned penteconter of split ends, and the curvature of the central part is
minimal. We can deduce, then, that in the temple, the primary curvature is that which affects
the entablature and the roof equally, and the curvature of the stylobate is merely a
consequence of the first. Of course, the latter does not have, as has been claimed, the purpose
of evacuating rainwater. The inclined plane that results from raising a few centimetres to
dozens of metres is incapable of evacuating anything and furthermore, the rain in Greece is
scarce. The curvature of the entablature is a sophistication, not aesthetic but rather of the
mimesis, of the representation of the visual appearance of its model. The curvature was not
made so that it could not be seen from a distance, as it is commonly said; it was made and, as
occurs with all details, from a distance it just could not be seen.
Figure 18. Sheer forward and aft of a modern hull. Line drawing by the author, from a drawing by
Rémy Kaupp. <https://commons.wikimedia.org/wiki/File:Arrufo.svg>. License Creative Commons By-
Sa 4.0 International.
5. Ionic entablature
Vitruvius (4.1) attributes the origin of Ionic order to the incorporation by Ionian architects of
feminine traits into temples of originally Doric and thus more ‘male’ design. It is not a very
convincing explanation. Regarding Ionic entablature (I will discuss the columns later), the
peculiarities of this order can be explained by a variation—so far only hypothetical—in the
design of the gunwale of the penteconters of this region. Ionia in the 6th century BC came
under Persian control, like the coastal regions of Asia Minor and the Levant (Caria, Lycia,
Cilicia, Cyprus, Phoenicia, etc) and could develop traditions of naval construction both
imported from those nations and native, but different from those on the European continent.
In a penteconter whose structural strength in the face of hogging and sagging has been
assured by solutions such as reducing the proportion between length and width and placing
the oarsmen on two levels, like the dikrotoi penteconters in figure 14, the gunwale could have
been lighter. In that case, the gunwale would be joined to the hull with lighter struts and more
Sheer forwardSheer aft
separated than those struts that gave origin to the Doric triglyphs (figure 19 b). At the moment
of overturning that penteconter, the struts would not support the weight of the ship and would
break. There would be two solutions to this, which explains the two arrangements of the Ionic
entablature that we know of. The first would consist of sawing the struts and supporting the
hull directly on the gunwale eliminating the oarsmen’s windows (figure 19 d). In the resulting
entablature the dentils would rest directly on the architrave and there would be no frieze, as
was habitual in the Ionic temples of Asia Minor, such as the Temple of Zeus at Priene.
To understand the second option, we must remember that in the triremes of the
Classical period, in which oarsmen in the upper rows were very exposed to the elements (see
the famous Lenormant Relief), in combat the παραρρύµατα were deployed on both gunwales.
These were protective curtains of hide or sailcloth. The alternative solution, at the moment
of creating the entablature, would consist of representing the gunwales with the deployed
pararrymata, which would give rise to a continuous frieze which could be left smooth or
sculpted and painted with scenes representing the warriors who rowed behind them, as had
been done in the metopes of the Doric frieze (figure 19 e).
As for the three fasciae or bands of the architrave in Ionic order, they may correspond
to a gunwale of a tapering thickness, as represented in figure 19 a, or to a clinker built
Casson (1995, ch. 11 n. 99).
Clinker built: a method of construction in which the edges of planks overlap each other. En
Figure 19. Hypothetical board of an Ionic penteconter, at left (a and b). At center, the inverted board
(c), and at right two ways of petriﬁcation (d and e). Line drawing by the author. License Creative
Commons BY-NC-SA 4.0 International.
6. The temple plan
Scholars agree that the stone temples from the 7th century BC and later are the continuation
of previous constructions from the Early Iron Age (1100-700 BC), excavated in various
locations in Greece. Little is known of these other than the floor plan: Thermon A and B,
Lefkandi, Nikoleika, Eretria, Kalapodi, Ano Mazaraki and others. Some have a basilica floor
plan: long and narrow with the entrance at one end, which is in antis, with the other end being
semicircular or elliptical. Furthermore, their measurements range, with the odd exception,
from 20 to 35 m long and from 4 to 7 m wide (table 1). These characteristics are consistent
with the hypothesis that these structures are the supports of a complete overturned ship or of
part of it.
Before seeing an example of these structures in detail, it is necessary to make clear
that the penteconter was not the only warship in use during the Geometric period. There were
also the eikosoros, the triakontoros, and the tessarakontoros, of 20, 30, and 40 oarsmen
respectively, of which the most common was the triakontoros. The version of this vessel that
had one row of oarsmen measured 23 m in length, with the two row version measuring 13.5
m. As for the penteconter, in addition to the single level type used for troop transport (38 x
3.9 m, a proportion of 10:1), there was a two-level version and a penteconter merchant
variety. The latter was wider to accommodate cargo and had a proportion of up to 5.5:1, that
is, a beam that could reach 7 m. This variety of measurements is contained in a range of 20 to
40 m in length and 4 to 7 m in width, which fits nicely with the size of the excavated
buildings. To compare ships with buildings, I show in figure 20 the floor plan of a penteconter
merchant ship of those measurements: its shape is tapered at both ends and has at least two
main beams in the union of the bow and stern with the central body, and banks of oars that
act as secondary beams, even though some may have been stronger, acting as beams in a strict
sense; entry to the ship was through the stern where there was often an aphlaston or similar
Figure 20. Plan of a hypothetical merchant penteconter with a ratio of length to beam of 6:1. Drawing
by the author. License Creative Commons BY-NC-SA 4.0 International.
Table 1. Plans of some buildings usually considered predecessors of Archaic stone temples.!
Line drawings by the author.
Building U2 (Tarsus)
Casson 1995, 56.
Ratio between 5.5 and 6.5 to 1 (Casson 1995, 158).
Casson 1995, 46.
Megaron A (Thermon)
22 x 6
9th century BC
47 x 10
8th century BC
¿20? x 7
35 x 7.5
24.6 x 7.6
7th century BC
27.9 x 7.5
Building T (Tiryns)
20.9 x 6.9
Megaron B (Thermon)
21.4 x 7.3
8th century BC
18 x 6.5
8th century BC
33 x 6.5
Temple of Apollo
The Megaron A of Thermon (table 1), dating roughly from between 1100 and 700
BC, has an approximate proportion of 4:1. It has a tapered end, corresponding to a bow, and
an entrance end in antis, which corresponds to a stern that is cantilevered or has been
sectioned to avoid the aphlaston or similar structure making access through the door difficult.
The Megaron also has the internal division that results from projecting on the floor the
structure of the penteconter merchant ship: pronaos, naos, and opisthodomos, that will be
traditional in stone temples. Curiously, in the naos, overturned pythoi and amphorae were
found containing ashes and animal bones; the position of the pythoi and amphorae could
have obeyed a desire for coherence with the ceiling of the building, wich—in our opinion—
consisted in an overturned ship.
The Lefkandi megaron from the 9th century BC is very similar but much bigger (47 x
10 m). Similarly long were, with one apsidated end and the other straight (although without
a division of the interior space), the building at Nikoleika, the Hekatompedon at Eretria and
the south temple at Kalapodi. The building at Ano Mazaraki has a similar floor plan, with a
single very significant peculiarity: the entrance end is also apsidated, like the stern of a ship.
Mazarakis Ainian 1997, 44.
Mazarakis Ainian 1997, 45.
The larger dimensions of Lefkandi may correspond to a ship that was not Greek, but rather
Egyptian, or constructed according to Egyptian tradition. The Middle Kingdom poem The Tale of
the Shipwrecked Sailor (2025-1710 BC) mentions a galley of 63 x 21 m: “I had gone down to the
sea, / in a boat 120 cubits long and forty cubits broad, / in which there were 120 sailors from the
choicest of Egypt” (Abbas 2013, 11). Archaeological evidence reveals commercial exchanges
between Lefkandi and Egypt which may have been direct. Homer (Il. 2.509-510) claims that the
Boeotians went to war with Troy in ships transporting 120 men, and Lefkandi is barely 5 km
from the Boeotian port of Aulis, from which, according to legend, the Greeks departed for Troy.
All of these buildings as well as Megaron B of Thermon and the U2 building at
Tarsus also shared the following architectural peculiarity: the long walls are not completely
straight, but slightly convex, which can be easily explained as the translation to the building
perimeter of the form of the hull of a large ship, which is not exactly straight in its central
section, but slightly curved for reasons of hydrodynamics.
Other long buildings from similar periods are straight at both ends, as if the bow, in
the same way as the stern, were cantilevered or, more probably, sectioned. These are Building
T at Tyrins, the Megaron B at Thermon, the temple of Apollo at Halieis, the Megaron Hall at
Emporio and the first Heraion of Samos. Lastly, there are buildings of similar width but
shorter, such as the apse-shaped building at Nichoria (13 x 8 m) that may correspond to a ship
cut in half. If ships captured in battle were dedicated to the gods, there is the possibility that
the captured hull had been badly damaged by the effect of the spur, and that an important part
of its length had remained unrecoverable.
Many of these buildings have a line of central columns in contrast with the double
lines that would be common in later periods. This arrangement of the columns is logical in a
building with a gable roof, but in the case of an overturned ship this arrangement is even more
justified. A ship’s hull is designed to support the force exerted by the water on both sides of
the ship upwards and towards the center of the structure. To counter this force, there are
transversal beams in the upper part of the hull, subjected to compression, or rowing benches
that, if they cross the structure, fulfil the same function (figure 21 left). However, in an
overturned hull that rests the bulwark (the upper line of the gunwale) on lateral wall supports,
the weight of the keel and hull pushes the hull down and outwards. In this position the beams
Mazarakis Ainian 1997, 57-58.
Kyrieleis 1993, 178.
have no effect because they haven’t been designed to be subjected to the opposite force,
traction, and they fall, the whole structure warps and breaks (figure 21 center). The only way
to avoid lateral pressure on the walls is to translate the weight of the keel vertically towards to
floor, with a central line of columns for example (figure 21 right).
Figure 21. Overturned boat without (centre) and with central support (right). Line drawing by the
author. License Creative Commons BY-NC-SA 4.0 International.
The so called Double Stoa at Thorikos, from the last quarter of the 5th century BC,
was according to all the evidence, covered by an entablature and a roof of wood, in the
ancient manner. Its dimensions (31.9 x 14.7 m) make it improbable that it had the hull of a
real ship for a roof, but it is likely that it copied in stone and of a greater size— as did so
many other buildings from the period—an older building of similar characteristics. Therefore,
the architect did not think that a central line of columns was stable enough, and came up with
a surprising solution: a solid wall of stone the length of the central axis, over which the keel
Onians explains the axial colonnade as a “subliminal association with a trirreme”: “The single
column at the rear, as in the earliest temple at Samos with its row of posts down the center, gave
the temple plan an axis much as the keel and stern did a ship” (2002, 54).
Miles 2015, 171-172.
would originally be placed, with a door between the two halves of the covered space (figure
Figure 22. Plan of the Double Stoa at Thorikos (after Petrakos 1997 and Miller 2015). Line drawing by
the author. License Creative Commons BY-NC-SA 4.0 International.
7. Bows, adyton and thesauroi
In addition to the longer buildings that we have just analyzed, there were others that were
significantly shorter, of which the most characteristic example is the Daphnephoreion at
Eretria, from the middle of the 8th century BC (9.75 x 6.5m). Its apsidated form suggests that
it was the result of cutting and overturning only the bow of a penteconter, which would
convert it into the equivalent of the opisthodomos of larger temples.
It is well-known that certain Greek temples were not only sacred places but also
repositories of the city’s treasure, in the form of valuable works of art, metallic and otherwise,
as well as cash. These works of art were stored in the adyton, the back room known as the
opisthodomos, in other words immediately below the overturned bow in which they were
The Corcyrean Portico of Elis had a similar distribution. The first version of this building, from the
end of the 5th century BC, was built from the spoils that the inhabitants of Elis took from
Corcyreans who had “reached their lands by ship” (Paus. 6.24.4).
usually transported. In later temples the wall of the opisthodomos opened so that its contents
could be seen from outside through a lattice, creating a porch in antis parallel to the pronaos.
In this case, the door to the cella disappeared, making the bulkhead between the bow and the
rest of the hull project over the ground without interruption.
Written and material sources bear witness to the offering of ships in sanctuaries such
as Istmia, Sunio and Samos. In the sanctuary at Olympia the cities dedicated small temples,
known as θησαυροί, aligned in rows on a platform contiguous with the temples of Hera and
Zeus. These treasures are similar in size, shape and function to the adyton of the larger
temples and for that reason it is fair to interpret them as buildings created to support the
overturned bow of a ship to store and exhibit the valuable contents that were transported in its
interior. The sanctuary at Olympia had a unique “flagship”, the temple of Zeus, and what we
can interpret as bows of various cities that competed to lead this flagship. They did so by
exhibiting wealth that acted as a visible indicator of their economic and military success. In
the sanctuary at Delfos something similar occurred: there was a single temple dedicated to
Apollo, and various treasures in the form of a small temple, but there were also non-
architectural treasures that were instead sculptural, consisting of bronze sculptures. Herodotus
(8.121) claims that with the firstfruits of the Battle of Salamis the Greeks dedicated in Delfos
a statue in bronze that held a ship’s ram in its hand. It appears logical to think that these
firstfruits consisted of the bronze rams of captured ships, recast to create the statue that in a
very didactic way shows in its hand the object from which it has been made. Herodotus also
informs us (3.59) that on another occasion the Aeginians offered in the Temple of Athena at
Aegina the rams of ships captured from the Samians. The written and material sources
The more ancient of the two was originally a temple of Zeus, and only later when the new temple of
Zeus was built did the older one change its invocation and become dedicated to Hera.
therefore enable us to reconstruct a hierarchy in the dedication of ships in the sanctuaries that
in descending order consist of: a) a complete ship, b) a bow, or c) a ram, the last usually
8. Peristasis as enlargement
At the present time, due to the construction of certain temples on the remains of others, and to
the lack of rigour in the primary excavations, it is difficult to know what the first temple to
have a peristasis was. In the opinion of the most sceptical scholars there is no certainty that
any temple from the 7th century BC had it. However, I believe that the origin of the
peristasis can be explained independently of the date and place in which it arose, attending
only to its shape in the temples of the sixth and fifth centuries BC.
If the explanation of the origin of the entablature and Doric and Ionic friezes that has
been given here in the previous paragraphs is true, the presence of a frieze at the top of the
cella in temples with peristasis supposes a difficulty: if the frieze depicted an overturned
ship’s gunwale, a building with two friezes, an inner one over the cella and an outer one over
the epistylium, would it have been done with two ships?
In architecture, the usual thing is to first construct the walls and later the roof, its
dimensions in agreement with the space it is intended to cover. However, in the construction
of walls that are to support an overturned ship, the order is inverted: the ship was built earlier
with measurements required for its nautical use. The walls were built later, to fit the ship that
already existed. This way of proceeding inevitably led to both medium and long term
problems: the passing of time and the action of the elements would end up ruining the ship’s
hull making it necessary to replace it. In order to avoid demolishing and rebuilding the
Wilson Jones 2014, 48-53.
supporting walls with new measurements, builders would have to obtain a ship with the same
dimensions as the previous one, and that was not easy, even more so in the case of ships
captured from the enemy. If the replacement was to be done with a bigger ship, an extension
was required, preferably preserving the existing building, either for reasons of economy (to
avoid having to construct a new cella) or for conservatism. The enlargement of the
measurements of the new roof beyond the dimensions of the existing supporting edifice
would explain the presence of peristyle, the columns of which supported the gunwale of the
new overhanging ship, giving rise to a peripteral gallery that, incidentally, has no known
practical function (figure 23).
Furthermore, the respect for the remains of the previous building would explain the
presence of the frieze above the cella, which would correspond to the gunwale of the original
ship, which has been preserved despite the fact that the rest of the hull had been ruined.
Therefore, the answer to the previous question is effectively this: a temple with peristasis
contains the remains of two ships: an older, smaller one above the cella, and a larger, more
modern one covering and overhanging the original on all four sides.
Figure 23. Creation of the peristasis as a process of replacing a ship with a bigger one (from left to
right). Line drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
On the penteconter there were only oarsmen’s windows port and starboard, not in the
prow or stern; therefore there should only have been windows in the long sides of the temple.
But in temples with peristasis the gunwale goes round the entire building, including the
shorter sides. This obliges us to imagine a ship longer than the temple, of which have been
preserved the entire length of the gunwales, but only part of the hull, as shown in figure 24.
The initial and final parts of the port and starboard gunwales were later folded inwards in
order to enclose the two shorter sides of the building, meeting in the center and creating the
base of the pediment. I explained previously the slight curvature of the temple’s longer sides,
attributing it to the sheer of the length of the overturned ship, but the similar curvature of the
shorter sides was left unexplained. Ships do not have anything similar to transversal sheer,
but the curvature of the shorter sides of the peripteral temple can be explained by the fact that
this entablature originally formed the initial and final part of the longer sides and was affected
by the same curvature as those.
Conversely, the beams and the deck, when present, are slightly convex, more elevated in the center
than on the sides, to allow the bowing of the structure in response to the force exerted by the
pressure of the water from both sides towards the center of the structure.
Figure 24. Hypothetical section of bow and stern, and folding of the fore and aft ends of the gunwale.
Line drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
The evolution of the proportions of the temple plan may be summarised in this
simplified way (figure 25). Starting with a merchant penteconter with a proportion of 6:1, by
eliminating the bow and stern sections we get a cella with a proportion of 4:1. A ship three
times the size is superimposed on this, with gunwales of a length nine times the width of the
first ship. Then, the protruding parts of the gunwales are folded to skirt the entire peristasis,
covering a total space of 6:3, or 2:1, the approximate proportions of Classical temples; in the
case of the Parthenon 2.2:1 (69.5 x 30.9 m).
Figure 25. Schematic proportions of the cella and the peristasis in the Classical period. Line drawing
by the author. License Creative Commons BY-NC-SA 4.0 International.
There is debate among archaeologists as to whether some of the more ancient
buildings, such as the Megaron B of Thermon (9th century BC) or the Hekatompedon at
Samos (8th century BC) were enlarged with a peristasis while the interior edifice, the cella,
was still standing. Be that as it may, the truth is that most later stone temples with peristasis
were not built in two stages, but rather in only one that included an interior cella and
peristasis from the start. This apparently rejects the idea of enlargement. But the enlargement
could have been real or merely simulated, as a way of showing the age of the building. Many
temples were built on an emplacement in which there had been similar previous
constructions; the Parthenon was probably the third of a series of similar temples, and the
temple of Artemis at Hyampolis, present-day Kalapodi, the fifth. The construction of
temples was furthermore the initiative of the ruling class and a citizenry who, as always, had
among their legitimizing arguments their age-old lineage and social position. One way of
exhibiting this ancient heritage consisted of showing the antiquity of their worshipping of the
gods, manifested in the succession of temples built by them and their ancestors.
Spawforth 2006, 50.
The suitability of the idea of enlargement is confirmed by one more piece of evidence:
the plans of several temples contain up to three stages of enlargement. In the great open patio
of the colossal temple of Apollo at Didyma there is a ναΐσκος whose plan is identical to that of
a small, simple archaic temple; the term naiskos is thought to be a diminutive of naοs and is
translated as ‘small temple’, but it could be of naus and be translated as ‘small boat’. The
enormous Temple of Apollo, or Temple G, at Selinunte has a plant that has been interpreted as
a combination of covered spaces and courtyards, but can be better interpreted as a succession
of three temples, from smaller and older to larger and more recent (figure 26). The peristasis
(in black) and the enlargement of the cella (in green) are standard; what is notable about this
temple is that inside the cella the remains of an older structure (in red) can still be identified, a
structure with the characteristics of the most primitive naval buildings. This structure would
give support to an overturned ship, and would have a room closed to the west, equivalent to
the adyton and destined to store the wealth won in battle, which would underpin the bow, and
a central space delineated by the columns which would support the larger part of the ship,
which originally would be enclosed with the ship’s sails and other material and would provide
temporary shelter for the crew; the entrance and the stern, as usual, would be on the eastern
side. The whole temple was constructed all at once, but probably represented a history of
occupation from the pioneers to the first colonists, and from these to the contemporary
inhabitants of Selinunte, and affirms therefore the antiquity and consequent legitimacy of the
presence of the Greeks in this land. Lastly, as an example of archaeologically-evidenced
enlargement, albeit from a later period, there is the Temple of Apollo at the Letoon of Xanthos
(2nd century BC) which preserved in its cella a smaller and older Lycian temple from the 4th
Figure 26. Temple of Apolo G (Selinunte), whose design simulates a construction in three phases (red,
green and black). Line drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
9. Number and position of the columns in peripteral temples
Vitruvius associates the columns of a temple with a human being (4.1.6-7), in concordance
with the architects who gave to the columns of certain buildings the form of Atlas (Temple of
Olympian Zeus in Agrigento) or Caryatids (Erechtheion in Athens). If we interpret the number
and position of the columns in light of this association of column and person, we obtain a very
The penteconter had, originally, a crew of fifty people. But we must not be confused
by the etymology of the term πεντηκόντορος ‘with fifty oarsmen’; there could have been
more, distributed in a different way. In Homer, the penteconter of the Phaeacians is handled
by fifty oarsmen and two officials (Od. 8.34–35 y 48), surely the pilot and captain. The crew
of Odysseus’ ship, however, numbered 58 men: six are lost with the Cicones (Od. 9.60),
another six with Polyphemus (Od. 9.289, 311, 344) and forty six make it to the island of Circe
The pilot would carry out the function of helmsman from the stern while the captain would lead
from the prow giving orders to the oarsmen.
(Od. 10.203-209). Jason’s ship Argo has 55 crew-members in Apollonius of Rhodes’ version
Additionally, the penteconter was a warship, and the distribution that we usually
assume, with fifty oarsmen, was the appropriate one for a peaceful voyage in which it was
important to reach maximum speed. When the ship was on alert or in combat, however, a part
of the crew would act as marines (archer, slinger, or hoplite) while the rest would continue
rowing. We can calculate the approximate number of rowers and marines on board a
penteconter using three indicators; (a) the best preserved ship in the fresco of the Ship
Procession from Thera (figure 27), from the 16th century BC, has 42 paddlers (21 in the
visible side) and 14 non-paddlers, which gives us a crew of 56 people; (b) the penteconters of
the fleet created by Polycrates of Samos in the 6th century BC had ten archers (Hdt. 3.39.3);
(c) lastly, we know from Thucydides that the Athenian trireme of his time, which succeeded
the penteconter, was endowed with ten hoplites (Thuc. 3.95.2), to which were added,
exceptionally, four archers (Thuc. 2.23.2). The numbers are consistent, therefore: the total of
40 oarsmen, 10 soldiers (who, depending on the occasion, also rowed), plus a few sailors and
two officials, makes a total of between fifty and sixty people. Finally, the distribution of these
onboard is very obvious: the 40 rowers worked to the sides of the ship, while the ten armed
men, the sailors and the officials occupied the central part.
Terminus ante quem (Pearson, Brewer and others 2018).
Figure 27. Fresco of the Ship Procession from Thera. Marinatos 1974. Drawing © Wachsmann (1995,
So, in the temples of the sixth and fifth centuries BC we see the same number and
arrangement of columns as crew-members of a penteconter. Figure 28 shows the plan of the
temple of Hera at Olympia: the peristyle has 40 columns, 20 on its southern side, which
correspond to the oarsmen of the port side (in green), and another 20 on the northern side,
which correspond to the rowers of the starboard side (in red). In the interior there are twenty
columns (four of them in the antas and the cella), corresponding to the soldiers, sailors and
officials (in violet). The numbers are very similar in the temple of Zeus adjacent to that of
Hera at Olympia and in others shown in table 2.
Figure 28. Plan of the temple of Hera at Olympia, and groups of columns representing port rowers
(red) and starbord rowers (green) at the perimeter, and soldiers, sailors and ofﬁcers (violet) at the
center. Line drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
Table 2. Number and distribution of columns in some Archaic and Classical temples.
The poet Alcaeus and others after him made famous the metaphor of the state as
ship, and the ruler as κυβερνήτης ‘pilot’. Regarding the interior columns, it is striking that
there is no differential sign in them that would indicate a different status for one or both of
them, which could be a way of representing the captain and the pilot. This egalitarianism
between the members of a group superior to the rowers, a group consisting of sailors, marines
and officers, is consistent with political use in the Archaic and Classical periods: the time of
aristocratic rulers who opposed the one-person rule of the basileis of the Mycenaean period
and of contemporary tyrants, and who governed the poleis through collegiate bodies of few
Around 575 BC in Samos the architects Rhoikos and Theodorus built a temple to Hera
that was the model for several Ionic temples of colossal proportions. It boasted 134 columns
Temple of Hera at Olympia
7th-6th century BC
Archaic temple of Poseidon at Istmia
6th century BC
Artemis at Corcyra
6th century BC
North Building at Samos
6th century BC
South Building at Samos
6th century BC
Temple of Apollo at Corinth
6th century BC
Temple of Apollo at Eretria
6th century BC
Temple of Zeus at Olympia
5th century BC
Temple of Hera at Paestum
5th century BC
Temple of Apollo at Bassae
5th century BC
Alcaeus D. 46A, L.-P. 326, P. Oxy. 1234 fr. 1, 7-14.
It may be that the ship's captain is represented by the priest (alive, of course), who conducts the
sacrifice and fulfills the role in that moment of provider of food, a role in which captains,
governors and business-people in general like to present themselves.
and was the first to have a dipteral peristasis (figure 29). McEwen was right in intuiting that
the dipteral peristasis was inspired by the double row of oarsmen of two-banked ships. The
literary sources say little about biremes, a new type of ship which doubled the crew of the
penteconter, because this model was soon superseded in turn by the trireme, with triple the
penteconter’s crew. This type of ship would become the standard during the following
centuries. Homer claims that the Boeotians were the only Greeks who went to war in Troy in
ships that transported 120 men (Il. 2.509–510), which Casson interprets as a reference to a
bireme. Be that as it may, the bireme had two overlapping rows of approximately 25
oarsmen on either side, with some 100 rowers in total; we do not know how many non-rowers
made up the rest of the bireme’s crew, but we do know the number for the triremes of the 4th
century BC: in addition to the captain or trierarch, there were six officers, ten sailors (five in
the prow and five in the stern) and fourteen soldiers, making a total of thirty-one non-
rowers. If we interpret the temple of Rhoikos and Theodorus and its dipteral peristasis
(figure 29) in the same way that we interpret the temple of Hera at Olympia (figure 28), again
the number and location of the bireme’s crew and the columns of the temple coincide a great
deal. The dipteral peristasis has 104 columns that can be divided into two sides, port and
starboard (in red and green respectively), with two rows of about 26 columns on each side,
which approaches the number of oarsmen of each row of a bireme. In reality each side has
28 columns in the outer row and 24 in the inner row, due to the existence of corners, which a
McEwen 1993, 102.
Casson 1995, 59.
Morrison 1995, 64.
The number of rowers in each row of the bireme would not always be 25. Athenian triremes of the
4th century BC had four more rowers in the upper row than in the middle and lower rows: 31
thranite rowers, 27 zygian, and 27 thalamian respectively (Morrison 1995, 64).
ship doesn’t have. As for the cella and pronaos, their 30 columns (20 and 10 respectively),
coincide exactly with the trireme’s thirty-strong crew of 20 sailors and officers (in the cella, in
violet) and 10 soldiers (in the pronaos, in light blue).
Figure 29. Plan of the Temple of Rhoikos at Samos, and distribution of columns by its nautical
correspondence: 52 port rowers (red), 52 starbord rowers (green), 20 sailors and ofﬁcers (violet) and
10 archers (light blue). Line drawing by the author. License Creative Commons BY-NC-SA 4.0
The temple of Rhoikos and Theodorus and many of their imitators of similarly
colossal proportions were never finished. Naval construction of the Classical and Hellenistic
periods turned out warships of ever greater dimensions, culminating in the Tessarakonteres of
Ptolemy IV Philopator, a monster propelled by 4,000 oarsmen. However it must have been
obvious to the contemporary architects that the construction of temples could not advance in
parallel with the arms race that occurred at sea. Indeed, in the sanctuary of Hera at Samos
itself, the temple of Rhoikos and Theodorus was followed in time by the so-called North
Building and South Building, smaller temples with a simple peristasis and approaching fifty
columns, which—we may reasonably conclude—assumes a return to common sense:
representing the traditional penteconter of one single level, which would remain the standard
in successive architecture. It is worth pointing out a peculiarity in the South Building (figure
30): the peristasis follows the same pattern that is habitual in the cella, namely, that the stern
has been sectioned. It also has an odd number of columns, due to which—coinciding with the
political interpretation of the number and location of the columns—the odd center column in
the eastern side (in white) could well represent the captain at the prow end, as a figuration of
the tyrant that ruled Samos at that time.
Figure 30. Plan of the South Building at Samos, and distribution of columns by its nautical
correspondence: 19 port rowers (red), 19 starbord rowers (green), 9 sailors and ofﬁcers (violet), 5
archers (light blue) and possibly a captain at the bow (white). Line drawing by the author. License
Creative Commons BY-NC-SA 4.0 International.
10. Real and legendary ships
The explanation just given for the creation of the peristasis as the superimposition of a larger
ship on another of smaller dimensions, runs into an important difficulty. Even though warship
size did not stop growing throughout the Archaic, Classical and Hellenistic periods, their
measurements did not increase in the proportion that our explanation requires. The cella of
the temple of Poseidon at Isthmia from the 6th century BC, one of the first—if not the very
first—that we know with certainty to have peristasis, had a width of 5.75 m, which coincides
with the width of a penteconter, whether for war or commerce (of 4 to 7 m respectively). But
the combined width, including the peristasis, measures up to 14 m (figure 31), and in that
period in Greece ships of such beam were not built. The largest warship during the period of
construction of the temple of Poseidon was the trireme, but this type of ship, although it
doubled the tonnage of the penteconter and tripled the size of its crew, had practically the
same length and width as the penteconter: 37 x 5.5 m. It is true that non-oar propelled
merchant ships were wider and shorter than their military counterparts, but there can be little
doubt that ships represented in the temple were oared warships.
Figure 31. Preliminary restored section of Archaic Temple of Poseidon, by Frederick P. Hemans (1997)
(Gebhard 2001, ﬁg. 2). Courtesy of the American School of Classical Studies, ‘Excavations at Isthmia’.
The solution to this difficulty is suggested to us by a finding made in the sanctuary of
Hera at Samos. Archaeologists there unearthed a statue of an Archaic kouros, from the 6th
century BC and therefore contemporary with the first temples of the sanctuary that had
peristasis: the first Temple of Hera of Rhoikos and Theodorus and the North and South
Buildings. The Kouros of Samos, 5.25 m tall, would, with the missing feet, be three times the
height of a person (figure 32). Its excavator Kyrieleis says truthfully of its size that “the
gigantic size of such statues may be linked to the conception widely expressed in Greek
literature ever since Homer, that the heroes of earlier times were much larger than the ‘people
alive today’”. In the Archaic temple of Poseidon at Isthmia, the temple’s 14 m of total width
almost triples the 5.25m of the cella. It seems reasonable, therefore, to establish a link
between the Samos statue and others of similar size and the enlargements of contemporary
temples, and claim that the architects of the 6th century BC represented, at least with regard to
the peristasis, the ships of the heroic period. This fits perfectly with the heroic character of the
figures and actions depicted in the friezes: the labours of Heracles, Theseus, Perseus, Lapiths
and Centaurs, etc. It would be, then, a historicist architecture, which assumed an antiquity that
it did not have.
Kyrieleis 1993, 120.
Figure 32. Kouros of Samos (Samos Archaeological Museum). Photo: Adam Carr. <https://
commons.wikimedia.org/wiki/File:Samoskouros.JPG> Public domain.
According to our hypothesis, the first apsidated buildings of the geometric age had
real ships for a ceiling, ships that really did sail the seas, but it is obvious that the roof of the
stone temples were never a ship, and that at some point the substitution of one for the other
took place. It is logical that the creation of the peristasis, due to the width of the new
construction, which tripled the width of real penteconters was not done using hulls of real
ships. The Archaic temple of Poseidon at Isthmia, which we understand to be first confirmed
case of peristasis (figure 31), had entablature and columns of wood, but it was also one of the
first to have a stone frieze and a tiled ceiling. This appears to be the moment of transition
between a mixed structure composed of walls that supported an overturned ship and the
temple avant la lettre. The gabled roof with straight overhangs, assumes a compromise on the
part of the architects, who were unable to erect what should have been a cross or barrel vault
to bridge a span of some 15 metres. As a result of this, only the gunwale of the second ship
would be preserved, that is, the entablature, as occurred with the original ship, of which only
the entablature at the top of the cella would be kept. The same compromise would explain the
substitution, in the colossal temples at Ionia, such as the temple of Apollo at Didyma and
similar, of the roof over the cella for an open-air patio.
Regarding the historicist interpretation, in most temples the exterior and interior
friezes depict heroic scenes of the same antiquity. For example, in that of Hephaestus at
Athens the exterior metopes depict the feats of Theseus and the interior Ionic frieze of the
pronaos shows Theseus fighting the Pallantides. But at least in one case, that of the Parthenon
at Athens, we can take the historicist interpretation to the extreme, perhaps even too far, and
assume that for its designers the ship represented in the cella was smaller and, as it was below
the more modern one, older. Hesiod in Works and Days (106-173) transmits the well-known
myth of the five ages of man: Golden, Silver, Bronze, Heroic and Iron Ages. The historical
Greeks who built the Parthenon considered themselves members of the Iron Age, and we have
already seen that in the exterior metopes of that temple they depicted figures from the Heroic
Age. The inner frieze, that depicts a procession prior to the carrying out of a sacrifice in
honour of the gods, should therefore correspond to a previous age. According to Hesiod,
Bronze Age men only dedicated themselves to war, and those of the Silver Age did not
worship the gods. Neither looks like a candidate to construct a temple. It seems more logical
that the Athenians of the period attributed its construction—figuratively, of course—to men of
the Golden Age, famed for their piety, among other virtues; this way, the frieze would
represent these making a procession. Many scholars argue that the inner frieze of the
Parthenon depicts the quadrennial procession of the Great Panathenaea. They are not
misguided, because the Athenians of the 5th century BC, proud of their domination over other
Greeks, surely imagined the Golden Age men as being the same as them in appearance and
sharing the same customs.
11. Minor plan issues
Regarding the crepidoma, its reason for being is found in the first ships that were overturned
so that the sailors could set up camp in provisional refuges. Even today youngsters, when they
set up their first tent, learn to choose a location that is slightly higher than the surrounding
terrain to avoid, in case it rains, water building up and flooding the tent. The slight elevation
provided by the crepidoma fulfils this function.
The columns of the peristasis of temples such as the Parthenon are slightly inclined
towards the interior, which has been considered, without better explanation, an optical
refinement (figure 17). If we turn our thoughts once more to the sailors camped in their
overturned ships, we realize that it is not very safe to sleep under a structure of various tonnes
of weight supported on provisional wooden struts. The most obvious danger is not the struts
breaking, but rather the structure moving slightly to the left or right and, once the position of
the plumbline is lost, collapsing to one side. The solution is obvious and we see it in present-
day tents: the lateral struts are usually oblique and inclined towards the interior in order to
counter the tendency of the structure to fall towards one side. In the temple this inclination is
minimal because in a well-balanced stone construction it does not have a structural function;
since it entails a difficulty for the builders, the sensible thing is to reduce it to its minimal
expression, enough for it to be perceptible and fulfil its function, which is only figurative, like
the faint relief of a vein on the hand of an athlete sculpted in stone.
The entrance ramp of some temples may be a reminder of the moment (more
imaginary than real) when for the first time following the construction of the crepidoma, a
wooden ship was pulled towards the platform and lifted onto the struts. Or perhaps of the
moment when the heavy sculpture of the god was placed in the temple.
The Western Doric temples of Sicily are peculiar in that they include stairwells. In the
Temple of Concordia of Agrigento they are found in the entrance to the cella (figure 33)
whereas in Temple G of Selinunte they are situated inside the cella (figure 26). Greek sailors
used their overturned ships when they were overseas, far from their homelands, on the coasts
of Sicily or in the south of the Italian peninsula. At the beginning of this article we referred
Shackleton’s Antarctic expedition in which the sailors camped under their boats. The
chronicle of the expedition recounts that half of them slept on the ground while the other half
slept higher up, in the interior of the overturned hull whose rowing benches had been
converted by the new position into the floor of a kind of attic. The stairwells of Western
temples may be a reminder of the founding era in which the ceiling of the building was the
dormitory of the first explorers. We know that the roofs of these Western temples had ὀπαί
‘openings’ for illumination; this is the same term for the openings in the hull below the
gunwale, almost at the water-line, for the oars of the lower row, and therefore had been
inspired by these.
In ancient Greek θάλαµος had, among others, these meanings: ‘the inner part of the house’,
‘bedroom’ and ‘the lowest part of the ship’ (Liddell, Scott & Jones 1940, s.v. θάλαµος AI, A2a
and AIII respectively). Probably the original meaning was ‘the lowest part of the ship’ and, after
overturning the hull, its interior became an attic and it continued being called so.
Spawforth 2006, 89.
Figure 33. Plan of the Temple of Concordia (Agrigento), with two stairwells at the entrance. Line
drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
The majority of Western temples have the door facing east, something that is not
strictly applicable to those of mainland Greece. The sailors of antiquity spent only the sailing
season overseas, from March to November approximately. On some occasions they were also
forced to spend what the Romans called mare clausum, ‘closed sea’, which was basically the
winter months. Regarding the controversial question of the orientation of the temple door, I
agree with what has been shown by many authors, among them Lucian of Samosata (De
Τὸ γὰρ τῆς τε ἡµέρας πρὸς τὸ κάλλιστον ἀποβλέπειν (κάλλιστον δὴ αὐτῆς καὶ
ποθεινότατον ἡ ἀρχὴ) καὶ τὸν ἥλιον ὑπερκύψαντα εὐθὺς ὑποδέχεσθαι καὶ τοῦ φωτὸς
ἐµπίπλασθαι ἐς κόρον ἀναπεπταµένων τῶν θυρῶν καθ᾽ ὃ καὶ τὰ ἱερὰ βλέποντα ἐποίουν οἱ
Archaeoastronomers identify the orientation of the temple door to a specific point in the heavens
which they calculate in seconds and degrees. However, the orientation of the door facing the
morning sun in winter for reasons of comfort is not so precise: the point of the horizon at which
the sun rises moves several degrees during the winter, as the sun moves several degrees during
the early hours of each day. Pernigotti (2021: 129) points out that Etruscan temples have a
partially similar orientation: their open side is oriented to the celestial arc between sunrise and
sunset during the winter solstice, in his opinion to receive as much light as possible inside the
temple all the day long.
It faces the fairest quarter of the day (for the fairest and loveliest is surely the beginning);
it welcomes in the sun when he first peeps up; light fills it to overflowing through the
wide-flung doors in the direction in which the ancient used to face their temples.
In Greece, where naval building was not used as a dwelling place, the orientation of
the door is not homogenous, and would be subject to more varied considerations. It could
face east to follow the custom used overseas, or to the point on the horizon where the sea
becomes visible, or to a nearby source of water, as if its constructors thought that the whole
building or the statue of the god could feel ‘water nostalgia’. The three temples of the
sanctuary of Letoon of Xanthos (Asia Minor) have the door facing south in the direction of
the sanctuary’s sacred spring. The Karthaia temple on Keos, in the Cyclades, also has the door
facing south, in this case to the sea. However, in the sanctuary of Samos all the temples and
similar buildings have a door facing the centerpiece of the sanctuary, its great altar. A final
observation about the question of the temples’ orientation: as the door to the building
corresponds to the stern of a ship, the orientation of the ship in reality is the opposite of that
which is usually given as the building’s own: a ship with the door facing east has a bow
pointing to the west.
When the roof was an overturned wooden ship, the walls of the cella were strictly
speaking the only “naval building”, and therefore the work of stonecutters rather than naval
carpenters. However, the ashlar masonry walls of classical temples copy the working methods
of naval carpentry. The use of clamps and dowels in the union of the ashlars is inspired by the
mortise and tenon joint, typical in joining the planks of the hull, and by the subtype known as
The editor and translator of the text (Harmon, trans., 1913) considers that καθ᾽ ὃ καὶ τὰ ἱερὰ
βλέποντα ἐποίουν οἱ παλαιοί is a medieval gloss, but it is not understood why Lucian, a 2nd
century AD Greek, did not have as well-founded opinion of ancient architecture as a medieval
a butterfly joint. The obsession with each block resting exactly on top of the lower ones
without there being the least groove coincides with the perfect fitting of the hull’s planks to
avoid water leaking in. Lastly, the sanding down given to the surface of the wall once the
fitting of the ashlars is finished is similar to the sanding that is applied to the finished hull in
order to ensure perfect hydrodynamics.
12. The roof, the original hull and the cargo
Logically, the wooden roofs of the first Geometric period structures are not preserved, but it is
possible to gain an idea of their shape thanks to a few preserved clay models. The Perachora
and Ano Mazaraki models (figures 34 a and b) are ideal for supporting this article’s
hypothesis. Their roofs are in the shape of an ogival vault, as is the hull of an overturned ship.
It is true that there are contemporary models of smooth slopes and with roof tiles that appear
to refute this (figures 34 c and d), but the truth is that our hypothesis does not deny that simple
buildings imitated others built with hulls of ships.
Figure 34. Four clay models of buildings of the Geometric period. Drawings: (a) Payne 1940, pl. 9b,
(b) Georg Herdt, (c) Oikonomos 1931, ﬁg.15 and (d) Ferri 1948 (cf. Wilson Jones 2014, ﬁg. 2.9).
We know the roofs of peripteral temples from the 6th century BC very well. According
to our hypothesis, the gabled pediment is the result of sectioning both the bow and stern from
the original ship’s hull. Although in a ship from that time both structures had a clearly
different form, in the temple they ended up being the same. The builders did not completely
eliminate the bow, however, because this had very significant elements from the naval point
of view, such as the ἔµβολος ‘ram’ and the προέµβολοι ‘secondary rams’. Over the pediment’s
three angles were the acroteria; strictly speaking, the acroteria are not adornments and
sculptures but rather the platform that supports them. The term ἀκρωτήριον ‘extreme’ is also
used for the ships’ rams (Hdt. 8.121) and indeed the verb ἀκρωτεριάζω means ‘to cut the
rams’ of the ships captured from the enemy (Hdt. 3.59 and Xen. Hell. 6.2.36). The main ram
was level with the water line of the enemy ship you wanted to attack; logic suggests that this
is the ram represented in the principal acroteria of the pediment’s superior angle. Moreover,
the ram, or secondary rams, higher up in the stem, are surely those that have been moved to
the two lateral angles in the temple. The adornments and sculptures over the central acroteria
and the two angular acroteria will be discussed later.
The pediment’s interior space represents the ship’s cavity. A large number of the
overseas voyages in the Bronze Age (the age of heroes according to the point of view of first
millennium BC Greeks) were made to obtain metals: mainly copper and tin with which to
make bronze. The most realistic thing would be to represent the ship’s hull full of ingots.
Indeed, the pediment of the Apotheosis of Heracles, belonging to a temple at the Acropolis
from c. 560 BC, represents in the trim of the top what may be a cargo of ingots of two distinct
types and shape (figure 35). These may have been arranged alternately to make better use of
the space. However, a pile of ingots doesn’t exactly create a very attractive artistic shape. In
the Archaic period, wealth was often accumulated not only in the form of money but also in
bronze objects, such as cauldrons and tripods, and later in the form of anthropomorphic
sculptures of gods and athletes, which due to their value were often objects of dedication in
sanctuaries. They were kept in the temples’ opisthodomos and thesauroi. The simplest
interpretation of the pediment group of sculptures is that it represents the cargo carried by the
penteconter, and that—however redundant the idea is to us—these stone statues represent
bronze statues understood as a store of value. The pediments are therefore an exhibition of the
city’s wealth and at the same time a manifestation of its piety.
Figure 35. Pediment of the Apotheosis of Heracles (New Acropolis Museum, Athens). Photo:
Tilemahos Efthimiadis, <https://www.ﬂickr.com/photos/64379474@N00/3668009586>. License
Creative Commons BY-SA 2.0.
At this point, it only remains to explain the origin of the main statue of worship, which
is actually very simple. In the most ancient single-room temples the statue of worship was at
the back of the room, the place that corresponds to the bow of an overturned ship. Ancient
ships carried their two bow anchors here, as well as what was known as the ἱερὰ ἄγκυρα
‘sacred anchor’ (Luc. Fugitivi 13), ‘last hope anchor’ in English. It weighed twice as much as
the others and was used in desperate situations much like pleading with the gods to save the
ship and the lives of its crew. Most people do not know just how many ships and their crew
were lost when storms pushed them against the rocks of the coastline; in this situation an
anchor or set of anchors that maintains the ship far from the coast could be the difference
between life and death.
But an understanding of this ships’ gear in all its complexity, as well as the related
questions of the form and meaning of Doric and Ionic columns and their capitals, will take us
far: to the use of artillery in naval warfare in the Aegean Sea of the third and second
millenniums BC. There is a problem, however: that chapter in the history of war at sea has not
been written yet.
13. War anchors and naval battles of the 3rd and 2nd millenniums BC
Scholars agree that before the invention of the ram, usually dated to the 10th century BC,
war at sea consisted simply of launching arrows, spears and slingshot projectiles from
distance, and once contact was made with the enemy ship, boarding it. However,
Thucydides tells us in a description of a skirmish in the port of Syracuse in 413 BC that
merchant ships dropped large dolphin-shaped weights of iron or lead from the ends of their
yards, qualified as delphinophoroi, puncturing the hulls of enemy ships:
αἱ δὲ τῶν Συρακοσίων νῆες µέχρι µὲν τῶν ὁλκάδων ἐπεδίωκον· ἔπειτα αὐτοὺς αἱ κεραῖαι
ὑπὲρ τῶν ἔσπλων αἱ ἀπὸ τῶν ὁλκάδων δελφινοφόροι ἠρµέναι ἐκώλυον. δύο δὲ νῆες τῶν
Συρακοσίων ἐπαιρόµεναι τῇ νίκῃ προσέµειξαν αὐτῶν ἐγγὺς καὶ διεφθάρησαν, καὶ ἡ ἑτέρα
αὐτοῖς ἀνδράσιν ἑάλω. (7.41.2-3)
The Syracusan ships pursued them as far as the merchantmen, where they were stopped
by the beams armed with dolphins suspended from those vessels over the passage. Two of
the Syracusan vessels went too near in the excitement of victory and were destroyed, one
of them being taken with its crew.
In Aristophanes there is a fleeting mention to this practice (Eq. 762). Similarly, a
painting on a Cypro-Archaic jug depicts what looks like a pyramidal stone anchor being
dropped from the yard of a ship (figure 36). More than a thousand years later, although more
Mark 2008 argues that the iconographic evidence provided so far for this dating actually represents
a wooden cutwater, and that the naval ram, sheathed in bronze, was invented in Greece in the 6th
or 5th century BC.
Coates 1995, 136.
Crawley, trans., 1910.
sophisticated weapons, such as ballistas and Greek fire siphons, had been introduced into
naval warfare, this simple resource was still in use in Greece. The Byzantine emperor Leo VI
reports that greatest dromons had a wooden castle in the centre of the bridge from which ‘ἢ
λίθους µυλικοὺς ἢ σίδηρα βαρέα, οἷον µαζία ξιφοειδῆ’ (‘or mill stones or heavy iron
[weights], like sword-shaped blooms’ ) were dropped, by means of a crane, onto enemy
warships to destroy them or, failing that, to kill some of the crew (Tactica 19.7 and 67).
Recently has been recovered from the Antikythera shipwreck what can be the first surviving
specimen of such a war dolphin.
Well then, the truth is that before the invention of the ram in the 10th century BC all
Greek ships, both merchant and military, had mast, yard and stone anchors and could
therefore use this form of primitive artillery, known as gravity artillery. So that it is reasonable
to investigate hereafter how long this type of weapon had been used.
Translation Elizabeth Jeffreys (Pryor and Jeffreys 2006, 487).
Marchant 2016, 463.
Ciordia 2007. An important part of the argument that follows was presented in “Del shadouf a las
columnas dórica y jónica” (‘From shaduf to Doric and Ionic columns’), a presentation given at
the XII Spanish Classical Studies Congress held in Valencia (Spain) in October 2007, which was
not published with the proceedings entitled Perfiles de Grecia y Roma (Madrid, 2009). I
published it on-line in Spanish with the same title and the URL <https://pompilos.org/
file_download/4/> (2007-10-25), but found no response.
Figure 36. Pyramidal stone anchor apparently thrown from a yard, painting on a Cypro-Archaic jug
(750-600 BC) (Nicosia, Cyprus Museum 1947/I-16/1). Drawing © Wachsmann (2019, 35 ﬁgure 34a).
In Mesopotamia, in the 3rd millennium BC, the well sweep, a primitive crane with a
counterweight at one end, began to be used to lift water from the river and irrigate fields; this
device was used in Egypt from the 2nd millennium, where it was called shaduf (figure 37),
and later in Greece. In Minoan Crete, the icon 25 of the Phaistos Disc depicts a ship with a
structure similar to the shaduf at one end (figure 38). The mechanism is similar to that formed
by a mast and a swinging yard: the yard of a ship from the time was fixed by ropes and
pulleys at the top of the mast and tipped to and fro like the arm of a weighing scale. Even
though was principally used as a crane for the loading and unloading of cargo, it is logical to
suppose that when needed its crew could drop the heavy stone counterweight, either loose or
attached to its shaft, on to an enemy ship in order to break its hull. Once this possibility is
accepted, we can go back several centuries to the Early Cycladic II and similarly interpret the
Yannopoulos et al. 2015, 5033-5034.
ships’ gear represented in what are known as the ‘Cycladic frying pans’ (figure 39). A shaft
with a large fish-shaped figure emerges from an elevated structure. This figure has been
interpreted as an ornament or course stabilizer, although it is more likely that it represents a
sculpted projectile with the same whale or fish-shaped figure as the dolphins of the Classical
period. The elevated structure appears robust, in order to support a heavy weight, and it is
inclined so that it is capable of raising the dolphin further and letting it drop on to an enemy
ship with devastating effect. Roman writers tell us of similar ship-borne engines, but it is not
known if they were used at the time of the birth of the Greek temple, and we will not take
their testimony into account.
More examples of the same gear in Coleman 1985, fig. 5.
Roberts 1987, 311.
Vegetius (Mil. 4.46.3-4) tells us of an iron-tipped naval ram, known as asser ‘bar’. Quintus Curtius
Rufus (4.3.24) describes the corvus ‘crow’ and the ferreae manus ‘hands of iron’, similar to the
throwable dolphin, but he function of this was not to puncture the hull but to grapple the enemy
ship in order to board it. Vegetius (Mil. 4.21.6-7) also reports the use of the tolleno ‘swape’, a
traction crane similar to the shaduf, which was used to lift soldiers to the top of a besieged wall.
Figure 37. Egyptian shaduf (Edwards 1890, 73).
Figure 38. Sign number 25 of the Phaistos Disc. Line drawing by the author. License Creative
Commons BY-NC-SA 4.0 International.
Figure 39. Cycladic oared ship incised on a ‘frying pan’ (2800–2300 BC). National Museum of Athens,
4974. Line drawing by the author. License Creative Commons BY-NC-SA 4.0 International.
Naval projectiles were not only whale or fish-shaped. The best preserved ship in the
fresco of the Ship Procession from Thera (figure 27), from the 16th century BC, has a
rampant feline figure at one end, which is thought to be a figurehead or similar ornament. At
the same time, the captain has in what we shall assume is his forecastle a large pole reckoned
to be a στῦλος, the long sceptre that identified the captains of ancient warships. It is fair to
suppose that the feline is in reality a projectile mounted on the stylos at the moment of
combat, falling from on high in order to penetrate the enemy ship’s hull. This is confirmed by
the discovery in Malia (Crete) of a sceptre’s finial in the mixed shape of panther rampant and
axe, dating from between 1800 and 1700 BC (figure 40), which is certainly a reduced scale
version of a dolphin used in sea warfare. To evoke the sea, the surface of the piece is covered
with a relief of swirls. Most probably the stone would not be released from the shaft as a
projectile, but instead would be dropped from one side or the other depending on the effect
sought: the axe part would be used to strike and cut the gunwale, whereas the fist and head of
the panther figure would be used like an awl to puncture the inner boards of the hull.
Figure 40. Finial handle of scepter, Malia (Archaeological Museum of Heraklion, 144889). Photo: Zde.
<https://commons.wikimedia.org/wiki/File:Final-handle of scepter, panther, axe, Malia, 1800-1700 BC,
AMH, 144889.jpg> License Creative Commons BY-SA 4.0 International.
In a ship such as the one in the fresco from Tera the naval combat would take the
following form. Inside his forecastle the captain would hold the throwable dolphin, with the
heavy weight of stone, lead or wood up high, the forecastle not being covered (figure 41). As
the ship approached, the enemy would launch a barrage of arrows, spears and slingshot
projectiles which the captain would withstand using the protection of the tower, the sides of
which had the shape of a great shield. When the ship was metres away from the enemy, one
side of the tower would open and the captain would move the stylos to drop the weight on to
the inside or side of the enemy hull. At that moment the captain would be even more exposed
to enemy projectiles, and made sense a piece of hitherto little-understood hardware.
Figure 41. Hypothetical representation of an ofﬁcer on the bow in bronze armour ﬂying a throwable
dolphin. Drawing © by Laura Ciordia, author’s daughter.
The Dendra armour (figure 42), from the end of the 15th century BC, formed by
several bronze plates, hardly allows the movement of the warrior in exchange for fully
protecting him against impacts. There is no trace of any δόρυ ‘spear’ with the armour, which
was the warrior’s weapon of choice, according to Homer. Everything mentioned so far allows
King 1970, 294.
to think that we are dealing with armour meant for naval combat using a throwable dolphin.
Late Roman war theorist Vegetius wrote centuries later:
De onere namque armorum nemo potest conqueri, qui stans pugnat in nauibus; scuta
quoque ualidiora propter ictus lapidum et ampliora sumuntur.
So must not complain of the weight of arms those who, as they fight on a ship, do not
move from the site; in addition the shields have to be stronger and wider to resist the
impact of the stones. (Veg. Mil. 4.44.4).
Figure 42. The Dendra armour (Archaeological Museum of Nafplio). Photo: C messier. <https://
commons.wikimedia.org/wiki/File:Πανοπλία των Δέντρων 8005.jpg> License Creative Commons BY-
SA 4.0 International.
Presumably throwable dolphins would take many other forms. The shaduf is literally a
huge scale, and we know of many hand held scales whose weights may well been reduced
versions of throwable dolphins. In the Uluburun shipwreck of the 14th century BC,
archaeologists found, in addition to weights of geometric shapes, 19 bronze zoomorphic
weights in the various shapes such as a sphinx, lion, lioness, bull, and duck. Literature also
provides evidence of what may be a throwable dolphin, this time in the form of a female dog.
Euripides describes in Hecuba (lines 1259-1273) the transformation of the Trojan queen into a
female dog with burning eyes which climbs a ship’s mast and throws itself into the sea, where
The sculpted dolphins were probably exceptional anyway. The usual thing would have
been to use as projectiles the weights from the ships’ usual cargo: stone anchors and in some
cargo ships, metal ingots. The ἐχῖνος ‘sea urchin’ of the Doric column gets its name from its
similarity to the sea urchin’s skeleton, and is at the same time very similar to the most
primitive typo of weight-anchor, which consisted of a flattened circular stone with a hole in
the centre. This was the standard type of anchor in the Mediterranean during the Bronze
Age, but it continued in use later, when pyramidal stone anchors and iron anchors were
invented. From the Assyrian siege of the Jewish city of Lachish in 701 BC, archaeologists
have recovered large flat pierced stones that the defenders hurled from the walls at the siege
machines (figure 43). These stones still contained remains of cords tied to the central hole
that the defenders used to recover them, making them similar to single-holed stone weight-
anchors. Indeed, the military theorist Vegetius showed that combat between two parallel ships
was similar to the siege of a wall (Mil. 4.44.1).
Pulak 1998, 209.
The type depicted in Frost 1991, 402, figures 31, 32 and 33, or in Kapitän 1984, 34, figure 2.2.
Ussishkin 2013, 15–17.
Also, an image on a Cypriot jug from the 7th century BC (figure 44), which is usually
interpreted as showing a sailor throwing an anchor, may be much better understood as a scene
of combat, with a circular anchor at the top of a stylos that has been split in half. The
expression of the character in the bow is neutral, whereas the one outside the stern on a rudder
is, shall we say, fearful. As we shall see later, this specific version of the throwable dolphin
with what looks like a stone circular anchor at the top has all the appearance of being the
origin of the Doric column and its echinus. Although several centuries later, the testimony of
Leo VI is very clear when he states, in the passage on the naval warfare already quoted
(Tactica 19.7 and 19.63), that λίθους µυλικούς ‘millstones’ are dropped from cranes on enemy
ships. It would be definitive if an ancient shipwreck were to yield evidence of the uses
described here. The truth however is that if anchors were used as projectiles in naval combat,
surely the archaeologists that found them would see anchors in the anchors, and if large
stones were used, they would interpret them as ballast or stones from the sea floor.
These circular anchors, hanging as was normal on both sides of the ship’s prow could be the origin
of the eyes, painted or in stone, found on many later ships. Used as projectiles, they may also be
the origin of the eyes that kill and petrify like those of the Gorgons, and the later basilisks. This
belief could be the result of a semantic slippage: “stone eyes that kill” could have become “eyes
that kill by turning into stone”. Lastly, circular anchors used as missiles may have been the origin
of the so-called ‘devil eye’; according to Apollonius of Rhodes (Argonautica IV 1669-1670)
Medea nullifies Talos's attack by bewitching him with her gaze. The evil eye was traditionally
counteracted by the digitus impudicus, the sign of the fig already attested by Aristophanes (Peace
546), that can represent a counter-attack with the ram.
The ship of Uluburun carried 24 stone anchors, all from a kind of sandstone and a type of Near
Eastern origin (Pulak 1998, 216). The number seems excessive for them to have been used as
anchors and they may have been taken on board as a reserve of defensive projectiles.
Figure 43. Perforated stones found at the foot of the city-wall of Laquis, thrown during the siege of 701
BC. Photo: © Ussishkin? (2013, ﬁg. 16).
Figure 44. Possible boarding and a throwable dolphin with broken shaft (usually interpreted as a
weight anchor attached to a rope). Cypro-Archaic jug from Cyprus (750-600 BC, British Museum
1926.6-28.9). Drawing courtesy by © Basch 1987, 261, ﬁg.564-B.
Zeus is often depicted as the god of the storm in the act of wielding a thunderbolt. In
the statuette in the figure 45 the thunderbolt is very reminiscent of a very simple type of
anchor, consistent in a bar with a central recess in which the rope was tied (figure 46). In
Roman iconography both sides of the thunderbolt often have a spiral surface, which may
correspond to the anchor line that has been coiled in spiral around both anchor arms. On
Greek coins such as the silver stater from Elis in figure 47, one arm of the thunderbolt has two
wings and the other what appears to be two Ionic volutes, whose meaning we will analyze
The simple bar anchor eventually become the stone stock of cruciform anchors, attached to the shaft
with a rope tied in the recess (Kapitän 1984, fig. 2.7–9). As the thunderbolt of the figure ?? is
isolated, it is logical to consider it a simple anchor, and not a detached piece of a more complex
If we force the comparison, we could say that the lower branches represent two tentacles of a
cephalopod like the squid, and that the whole thunderbolt is represented as an amphibious object,
which flies in the air as a projectile and dives into the sea as an anchor.
Figure 45. Zeus casting a thunderbolt, from Sanctuary of Zeus at Dodona, c. 470 BC
(Antikensammlung Berlin, Altes Museum, 10561). <http://media.web.britannica.com/eb-media/
Figure 46. Simple stone bar anchor with rope (left), and more accurate versions (right). Left: drawing ©
Kapitän, ﬁg. 2.6. Right: line drawing by the author according to Gianfrotta and Pomey, cfr. Pomey
Figure 47. Silver stater from Elis (obverse and reverse) from c.432–421 BC (BMFA 1202). <https://
commons.wikimedia.org/wiki/File:Greek_Silver_Stater_of_Olympia_(Elis).jpg> Photo: Exekias.
License Creative Commons BY 2.0 Generic.
As for the use of ingots as projectiles, copper and tin ingots in the shape of a flat bun
were common. Those found in the Uluburun wreck had a diameter of 22 cm. The sporting
events of later ages were training exercises for combat and one of these was particularly
relevant: the throwing of a bronze discus, did not appear to correspond to any type of warfare
on land. It did however make sense at sea, if launched from a merchant ship carrying ingots
cast in the shape of a discus.
There is an obvious functional similarity between the dolphin that acted by the force
of gravity and the bow spur of the later period. Pausanias (6.20.10-12) wrote that the exit door
for the horses in the Olympia hippodrome had the form of a ship’s bow and that over the spur
there was a bronze dolphin at the top of a pole that would be dropped to indicate the start of
Pulak 1998, 193 and figure 7.
Regarding the long-jump with legs held together, it seems ridiculous that this would ever be used on
land, but makes perfect sense when interpreted as a leap from the gunwale of one ship on to
another in a close boarding manoeuvre.
the race. Pollux confirms that it wasn’t a decorative dolphin but rather a projectile when he
claims that ὑπὲρ δὲ τὸ ἔµβολον δελφὶς ἵσταται, ὅταν ἡ ναῦς δελφινοφόρος ᾖ ‘above the spur
stands a dolphin when the ship is delphinophoros’(Onom. 1.85-86). Lastly, the etymology of
the term ἔµβολος ‘ram’ confirms this interpretation: ἐν means ‘inside’ and the root of βάλλω
‘throw’, not ‘insert’ or ‘push’, which is what one does with a spur.
Wilson Jones argues that there is a close relationship between the tripods and a part of
the Greek temple, which he specified in the frieze and triglyphs. In my opinion, the
relationship of cauldrons and their tripods with the temple is another one. A bronze cauldron
is a very heavy object, which sailors would carry on board for cooking once disembarked and,
as such a heavy object, it could be used if needed like the anchor, as a projectile fired to break
the hull of an enemy vessel. This would justify the consideration of the tripod as a palladium
carried on a ship, as Apollonius of Rhodes refers in Argonautica 4.526–536. The tripod is
represented on a column in its condition as palladium, and in the entablature as a valuable
object carried on the overturned ship, and which can be seen through the windows of the
14. Traces in heroic cicles
If a dolphin thrown by an official proved decisive, to the point of achieving a victory in the
Bronze Age naval battles, it would have been an individual feat capable of legitimizing the
pre-eminent position of the captain, admiral, or perhaps the king. And we can assume that
some trace would be left in the oral tradition that was collected by the writers of the historical
Wilson Jones 2002 and 2014, 177–189.
The identification of the triglyphs with tripod legs, made by Wilson Jones, is not explained in this
way. Perhaps the similarity has no particular significance, and it is the result of applying the
same ornamentation to two similar smooth vertical surfaces.
The bronze Dendra-style armour-clad officer on the bow, who threw stones from the
top of a pole or stylos, must be the origin of the mythical character Τάλως. According to the
story from Apollonius of Rhodes (5.1638-1688), Talos was a bronze automaton who went
round Crete three times a day, defending it by throwing huge stones at invading ships. He is
bleeding to death when Medea manages to open a vein in his heel, through which he loses his
ichor. The feet are the part of the body least protected by the Dendra armour. Talos was the
Cretan god of the sun (Hesychius, s.v. talos), but later came to be identified with Zeus, the
bolt thrower, in the figure of Zeus Talaios (Hesychius, s.v. Talaios).
The name Τάνταλος may be a composite of Τάν, the Cretan equivalent of Zeus, and
Talos, and therefore related to the previous character. Mythographers tell us that Zeus killed
Tantalus with a thunderbolt, which suggests that Zeus replaced both characters as the
principal god in the Cretan pantheon. For the rest, the two versions of Tantalus’ punishment in
the afterlife tell of a great stone weight. One version claims that a stone weight —or the
Mount Sipylus (Ant. Lib. Met. 36)— balanced above his head, threatened to fall (Pl. Cra.
395e). The other claims that he was in a lagoon below a fruit tree: when he knelt down to
drink, the water receded and when he lifted his hand to pick the fruit, the tree raised itself up
(Od. 11.582-592). In the former version there is a stone that can fall from above, as in
weighing scales or crane with counterweight. In the latter Tantalus represents the stone
counterweight hanging from one end of a crane (or from the mechanism formed by the mast
and yardarm), and—following the swinging movements of the ‘tree’, or arm of the crane—
falls without entering the water and rises without ever reaching the arm of the crane. In turn,
Niobe, Tantalus’ daughter, upon dying transforms into a rock that, according to one version
that associates her with water, forever cries (Apollod. Bibl. 3.5.6). In another version she flies,
like flies a throwable dolphin, back to her homeland (Ovid Met. 6.302-312).
The name of Talos may have something to do with the root ταλάσσαι ‘to support’,
because the mythical automaton lifts great stones with his bare hands. From the same root we
get τάλαντον, which means ‘scales’ and also a unit of weight equivalent to 27 kg. A
systematic study of the composite stone anchors of the ancient Mediterranean concludes that
there were stones of two average weights: the largest weighed between 140 and 180 kg, and
the smallest between 22 and 38 kg, i.e. a weight that is very close to a talent. Therefore the
name Ἄτλας, which consists of the negative prefix ἀ- and the root -τλα-, can mean ‘the one
without a talent’ or ‘the one without a support’. His name would express that the titan lacked a
round stone anchor to place as a kind of Doric capitel on a stylos!being used as a support for
an overturned ship; it is worth remembering that the term stylos, as well as ‘naval sceptre’,
means ‘column’ of a building.
Rowers the world over, when moving small boats, often turned them over and carried
them on their shoulders, literally taking the figure of atlantes. According to Hesiod (Theog.!
517) Atlas holds the οὐρανός ‘heaven’ which in the Homeric poems is imagined as a vault.
The hull of an overturned ship is concave like the celestial vault, and in the center of the
caissons of the classical temple’s ceiling were sculpted and painted golden stars, creating a
depiction of the night sky that would have much pleased the sailors, accustomed as they were
to studying it in their vigils. The mimesis, to use an ancient term, is twofold: the temple
ceiling represents a hull that in turn represents a star-studded sky.
Tóth 2002, 113-114. The study refers to stone anchors with holes to accommodate two wooden
fluke arms, rather than simple round anchors. There is no reason to believe that the weights of
both types were very different.
Kirk and Raven 1957, 10.
The atlantes were known in Rome as telamones (Vitr. 6.7.5). The Achaean hero
Telamonian Ajax stands out in the Iliad for his great stature and strength. He wields an axe
and enormous shield made from seven layers of ox-hide and one of bronze that is compared
with a tower (Il. 7.219-223). He injures Hector by throwing a stone in the battle for the ships,
a stone that is µυλοειδέϊ ‘like a millstone’ (Il. 7.270) and therefore circular and perforated like
a weight-anchor with a central hole and like the echinus of a Doric column. As the name of
his father, Telamon, con be formed from the same root of ταλάσσαι ‘to support’, Ajax is
probably another personification of the throwable dolphin-wielding captain, and his
prodigious shield a mixture of an ox-hide forecastle and bronze armour. That he wounds
Hector by throwing a stone in combat on foot, rather than naval, should not confuse us:
Homeric poems are constructed with formulae, in other words verses and sets of verses that
were created to be used in a particular context and continued to be used later in contexts either
a little or very different.
The strong man of Greek mythology par excellence is Heracles. If we imagine our
captain carrying his stylos in one hand and a feline shaped throwable dolphin on his back, the
resulting image is the standard iconography of Heracles with his club in one hand and the
Nemean Lion’s pelt on his head and back. His first labour consists of defeating the Nemean
Lion, which may have been metallic, because it is said that its skin was so hard that the lion
had to be skinned using its own claws (Theoc. Id. 25.274–279). The Nemean Lion would
from now on be Heracles’ throwable dolphin, his preferred weapon. He takes it to Eurystheus,
who takes fright even though it is an inert object, forbids Heracles to enter Mycenae and
encloses himself in a bronze pithos buried in the ground (Apollod. Bibl.!2.5.1). According to
Apollodorus (2.4.11) Heracles makes a club in Nemea, that is, at the same time as defeating
the lion. He uses a bow and slingshot as secondary weapons, as was common in naval
combat. Lastly, he is depicted in Archaic period ceramics as having very large eyes, which
may be a reminder of the circular bow anchors that appear to be the eyes of the ship and
which were thrown with the help of the stylos.
The last myth in which we shall find trace of the use of throwable dolphins involves
Oedipus, the Sphinx and King Laius. We know of Bronze Age weights in the shape of a
sphinx. The Sphinx at Thebes is represented mounted on a stylos (figure 48), a word that
designates both a freestanding column and a naval sceptre, and according to the myth, Sphinx
is the sister of the Nemean Lion (Hes. Theog. 326-327), which we have already identified as a
projectile. Oedipus’ relationship with her is peculiar: he must kill her in order to marry the
queen and become king, in essence overthrowing a previous monarch. Sphinx acts in this
sense as King Laius or in representation of him. In fact, she arrives in Thebes after his death,
and in one version of the myth is his daughter (Paus. 9.26.3). The name Laius is striking:
Λάϊος may be derived from λᾶας ‘stone’ , which is applied above all to the stones thrown by
warriors (Il. 7.268-270), from which there is a derivative with -ι- λαιαί ‘loom stones’. In this
case its meaning would be ‘the stone man’ o ‘the one who has or handles stones’, in reference
to the stone throwable dolphin in the shape of a sphinx. The way in which the Sphinx dies is
also meaningful: plunging from a high cliff (Diod. Sic. 4.64.4), like a falling projectile.
Oedipus, after defeating her, occupies Laius’ place both on the throne and in Jocasta’s bed.
Later, when he abandons the throne, Oedipus plucks out his eyes. Taking into account the
value we have attributed to the round stone anchors, as projectiles and as the symbolic eyes of
a warship, the plucking out of Oedipus’ eyes may represent the renouncement of both his use
of the throwable dolphin and usurpation of power.
In the Uluburun shipwreck for example. Pulak 1998, fig. 22.
Figure 48. Oedipus being questioned by the Sphinx, mounted on a stylos. Attic red-ﬁgured kylix by the
Oedipus Painter, 470-460 BC (Gregorian Etruscan Museum - Vatican Museums, inv. 16541). Photo:
Carole Raddato. <https://www.ﬂickr.com/photos/carolemage/9665213064/>. License Creative
Commons BY-SA 2.0.
Even though it takes us away a little from the main objective of this article, I would
like to continue pointing out that this interpretation of the figure of the Sphinx may provide
the key to understanding the etymology of one of the most leading Greek words: πόλις ‘city’
or ‘state’, which has a variant πτόλις with a -τ-. They are assumed to be two variants of the
same word, but Hesychius (s.v. πτόλις) defines the latter as: πτόλις· ναῦγος ἢ πόλις (‘ptolis:
shipwreck or city’). For this author πτόλις is only synonymous with πόλις in one of two
acceptations, but it is a different word when it means ‘shipwreck’. The Indo-European origin
of πόλις ‘hilltop’ is clear, since it has cognates in Sanskrit and Lithuanian. If we consider
Beekes 2010, s.v. πόλις.
that πτόλις was originally a different word from the previous one, its etymology is actually
very simple: ἡ πτό-λις would mean ‘winged lioness’, from the root πέτ-οµαι ‘to fly’ and the
noun λίς ‘lion’, and would be applied to the Sphinx. The derivation is so obvious that, as is
often the case, it would not have been proposed before because any semantical relation with
this mythical animal had not been seen. According to myth the Sphinx ‘dedicated itself to
piracy with a naval force’ (Paus. 9.26.2), and Hesychius’ gloss “ptolis: shipwreck”, which
until now has not been understood, confirms the relationship of the character with naval
combat. If the stone Sphinx on a stylos was a naval weapon that identified the king who lived
in a fortress located at the hilltop, it is possible that a statue of her was erected either at the
entrance to the fortress in a bas-relief, as could be the case of the Lion Gate of Mycenae, or on
a freestanding column, like the one of the Naxians at Delphi. The phonetic similarity between
πτόλις and πόλις and their semantic proximity (a symbol of royalty and ‘hilltop’ respectively)
could have caused them to be considered synonyms with the final meaning being ‘state’.
15. Anthropomorphic missiles and the Hymn to Dionysus
At some point throwable dolphins began to take human form in order to represent the gods,
leading to Archaic period sculpture in wood, stone and bronze of a variety of gods and kouroi
and korai. Statues of the gods now appear in sanctuaries at the top of free-standing columns,
in the place previously occupied by sphinxes or lions: hence the winged Victory of the Votive
Altar of Callimachus at the Acropolis. At this point it is inevitable that we must refer to
baetylism, the belief that particular stones, some meteors, were the seat of the gods. Among
them is the legendary Palladium, the image of Athena whose possession ensured that Troy
It is surely no coincidence, but rather the result of a long and tortuous history, that even today the
entrance to certain Greek fortresses is decorated with a winged lion. Here it is a symbol of Saint
Mark and Venice’s medieval and modern naval hegemony in the region.
was impregnable, or the xoanon of Athena in Athens. The word Παλλάδιον is a compound
from the root πάλλω ‘to brandish, sway a projectile’ before it is thrown.
The gods were also often associated with stones: Rhea substituted her newborn Zeus
with a stone that his father Cronus devoured. The term Τιτάνες derives from the reduplicate
verb τιταίνω ‘to stretch’. The cords from which hung large weights were stretched and
tautened, as were the titans, and the same verb τιταίνω is used with the meaning of ‘to weight,
use a balance’ (Il. 8.69), in the sense of hanging two weights, one at each end of a balance
arm. One of the clearest pieces of evidence for the use of anthropomorphic statues of the gods
as throwable dolphins is the Homeric Hymn VII to Dionysus, a new interpretation of which I
shall propose below.
The first verses of the hymn (lines 1-12) are thought to describe the kidnapping of the
young Dionysus by pirates. They are on board a ship that is approaching a headland. In the
eastern Mediterranean coastal temples were common in which anchors were dedicated to the
gods. As well as being a demonstration of piety, these temples were a reserve of anchors
which could be used by ships what had lost theirs in a storm. In the verses that describe the
supposed kidnapping, Dionysus offers no resistance, nor does he move or speak. This is
because the young Dionysus may in fact be a lead statue of a kouros. His hair (line 5) and
eyes (line 15) are κυάνειοι ‘dark’ or ‘azure’, that is to say, the color of lead. The ropes and
Apollonius of Rhodes narrates in Argonautica (I 955-960) how a ship reaches the coast and
changes a light anchor for a heavier one, and the first one ends up being dedicated to the goddess
A similar custom continues to be practiced today in mountain refuges, where food and firewood is
left for other mountaineers who may need it.
I believe that κυανός designates in Homer the lead, which recently cut has a bluish gray tone, or
any of its alloys (for example with antimony which hardens it) and not an enamel or lapis lazuli,
as the dictionaries say.
knots with which the pirates intend to tie him up with, as one would tie an anchor, fall to the
floor (lines 12-14). The pilot, expert in the ways of the sea, deduces that this is happening
because this lead statue is not a simple anchor but a god—that is, an anthropomorphic
throwable dolphin—and makes them see its disproportionate weight: …καρτερόν; οὐδὲ
φέρειν δύναταί µιν νηῦς εὐεργής ‘…strong that he is? Not even the well-built ship can carry
him’ (line 18). The wonders that are narrated in verses 35-43 reveal that they are about god
Dionysus. The youth’s metamorphosis into a lion (line 44) terrifies the pirates, who take
refuge in the stern with the pilot. The sacred anchor was carried in the prow, where
presumably the sailors had placed the statue. The image of the lion in the prow coincides
therefore with the location of the throwable feline in the fresco of the Ship Procession from
Thera (figure 27). Finally (lines 52-53) the pirates threw themselves into the water and turned
into dolphins. It is meaningful that they turned into dolphins, the shape that throwable dolphin
adopted in the Classical period. Two anchor stocks from imperial Rome found in the
Western Mediterranean are decorated with a lion’s head, which is uncommon and interpreted
as a reference to the myth of the metamorphosis of Dionysus. If this association is true, it
would support this interpretation of the statue of the god as anchor and projectile. Lastly, the
use of the mast and yardarm as a µηχανή ‘crane’ for dropping a god-shaped projectile onto an
Evelyn-White, trans., 1914.
The subordination of the dolphins to the lion of the prow might indicate a hierarchy of naval
projectiles by size or by the material from which they were made.
Gianfrotta 1997, 47.
enemy ship, and thus deciding a naval engagement, reminds us of the dramatic technique of
the deus ex machina ‘god from the crane’. Surely this similarity is no coincidence.
16. The Doric Column
As mentioned above, each temple column represents a penteconter’s crew member. In the
case of the Doric column, the crew member is represented by his weapon, a stylos crowned by
a round stone anchor. This is not a merely symbolic representation: he has left his weapon
supporting the temple in his place, unlike Atlas, who cannot do so because he lacks it. This is
why the term στῦλος means both ‘naval sceptre’ and ‘architectural column’. It is logical to
assume that Minoan and Mycenaean columns, so similar to the Doric, represented the same
Obviously the architectural columns are proportionally much thicker than their model,
the naval sceptre. The shaft of the column was originally made of wood. Pausanias (5.16.1)
claims that in his time there remained a wooden column in the opisthodomos of the temple of
Hera in Olympia, and the mural of the Tumba delle Leonesse in the Etruscan necropolis of
Tarquinia has columns with the shaft and abacus of an obviously brown color, distinct from
the capital’s yellow coloration, which may correspond to the stone or a gilded material.
The use of mechane in naval combats supports the possibility that this word has an etymological
connection with µάχη ‘fight’ and with µάχαιρα ‘knife’, which Chantraine discards (1968, s.v.
µάχοµαι) and Beekes sees as possible (2010, s.v. µάχοµαι). The connection can be extended to
κράδη, a similar mechanism used in comedy, which is in turn related to the Latin word cardo,
that designates one of the two arms of the groma and the celestial axis.
The original meaning may be ‘naval sceptre’, since it would be part of a very fertile lexical family
that includes στέλλω ‘to dispatch’, στόλος ‘fleet’, etc. However, it is the only term formed with
this root that designates an architectural element.
Sapirstein (2016, 575–577) believes that it was not one of the structural columns of the temple, but
he does not dismiss it completely.
Therefore, the striation may represent the imprint left in the wood of the shaft by the
carpenter’s adze. The entasis, the thickening of the shaft a third of the way up the column, has
been put down to aesthetic reasons, or a mechanism for correcting a supposed optical
distortion (Hero, Deff 135.14), but there is room for an explanation that depends less on
subjective perceptions. The stylos or shaft of a throwable dolphin, from a physics point of
view, is a class 3 lever: the lever’s fulcrum is the tip that remains immobile on the ground, the
resistance is the stone anchor which resists movement due to its inertial mass, and the effort is
the force applied by the warrior when holding the shaft (figure 41). The weak point at which
the shaft may break is where the effort is applied, and to avoid such a break the shaft must be
thicker in that area. The very name ἔντασις ‘tension, straining’ indicates that a force is applied
to a column, but it cannot correspond to the weight of the echinus or the entire building
because neither wood nor stone becomes deformed under pressure by thickening. As the
entasis is found a third of the way up the column and a man would hold the shaft at a height
of approximately 1.5 m, we may deduce that the shaft would measure around 4.5 m. Triremes
carried replacement oars of 4.5 m, although it is not fully known whether the blade was
removable or not. [Reference ??] Oars are class 1 levers and their weak point is where the oar
rests on the tholepin, which was located a third of way along the length of a trireme’s oar. It
is not unreasonable to suppose that this weapon was invented by sailors who, by necessity and
using the materials they had at hand, used oar shafts as propellants of round stone anchors,
like clubs or hypertrophied axes. Those oar shafts (or throwable dolphin shafts, if oars were
not used) would become thicker where the lever is weaker, and this thickening is represented
Morrison 1995, 64-65.
Stem set in the gunwale on which the oar rests and turns.
The Doric column has no base because at the moment of making camp under an
overturned ship, the supports would be driven into the ground for greater stability. We
explained above the reason for the inward inclination of the columns.
Below the echinus, in some capitals there is a crown of leaves, and in others there are
circular grooves both at the end of the shaft and the base of the echinus. The cause of the
grooves may be very simple: the temple is a re-used ship from which valuable metal parts,
such as nail heads may have been removed. In the case of the capitals, it was the crown of
leaves that was set in those grooves. This crown may be the Greeks’ habitual victory sign,
placed this time to the throwable dolphin, but Ionic and Corinthian capitals include other
floral elements, for which a further explanation is necessary.
If Heracles’ club represents a throwable dolphin and the same goes for Doric columns,
then the latter also represent clubs. Heracles’ club is named ῥόπαλον, which is usually
considered to be a derivative of ῥέπω ‘to descend’. This verb is used to refer to the downward
movement of the balance arm, but its etymology may be different. The term ῥόπαλον may
be a compound of ῥο-ή ‘pomegranate’ or ‘water lily’ and the root παλ- already mentioned of
πάλλω ‘to poise, sway a missile before it is thrown’. Its etymological meaning would be
‘stalk of pomegranate or water lily’. The latter, specifically the yellow water-lily (Nymphaea
lutea o Nuphar lutea) (figure 49), is a better candidate for being a metaphor for the war
The guttae have a blunt finial, according to some because they represent wooden pegs that have
come out of their position a little; they may also be metal nails whose heads have been removed
A crown should be worn on the head, not on the neck. The band around the end of the shaft below
the capital has the name ἀστράγαλος ‘vertebra’, especially relating to the neck (Il. 14.466), and
the place where the oar handle joins the blade is known precisely as αὐχήν ‘neck’.
Beekes 2010, s.v. ῥέπω.
Liddell, Scott & Jones 1940, s.v. pallō.
dolphin: it is an aquatic plant, its flower emerges from the water on top of a stalk and has,
upon flowering, a shape similar to that of a round stone anchor; if stone anchors were gilded,
like palladia, the analogy would be perfect. Two passages by Latin authors confirm the
association of the yellow water-lily with the Heracles’ club. Marcellus Empiricus writes:
‘There is a herb that in Greek is called nymphaea, in Latin Heracles’ club’ (33.63), and Pliny
the Elder: ‘The plant called nymphaea owes its name, they say, to a Nymph who died of
jealousy conceived on account of Heracles, for which reason it is also known as heracleon by
some. By other persons, again, it is called rhopalon, from the resemblance of its root to a
club’ (25.37). There is a further architectural parallel: Egyptian capitals were often crowned
with aquatic flora, the blue water-lily or blue lotus (Nymphaea caerulea), which was also used
as an ornamental motif in Greece. The preference for an aquatic plant of this color may be due
to the use of projectiles and anchors of lead, a bluish metal.
According to Aristophanes (Ach. 544-547): καὶ κάρτα µέντἂν εὐθέως καθείλκετε / τριακοσίας ναῦς,
ἦν δ᾽ ἂν ἡ πόλις πλέα / θορύβου στρατιωτῶν, περὶ τριηράρχου βοῆς, / µισθοῦ διδοµένου,
παλλαδίων χρυσουµένων… “Far from it, you would at once have sent three hundred vessels to
sea, and what an uproar there would have been through all the city! There ’tis a band of noisy
soldiery, here a brawl about the election of a Trierarch; elsewhere pay is being distributed, the
Pallas figure-heads are being regilded…” (Anonymous trans. 1928).
The identification of a mace-head with a flower is already present in the Assyrian iconography. See
the maces and bracelets of the warriors depicted on the reliefs of the Palace of Ashurnasirpal II
(Layard 1849, plate 5).
Bostock and Riley, trans. 1857.
There is a parallel of the use of pomegranates associated with capitals in the description of the
columns in the temple of Solomon in 2 Crónicas 4.12-13. It is claimed, moreover, that these
pomegranates are in nets.
Figure 49. A yellow water-lily, the ﬂower of the Nuphar lutea. Photo: Rosa-Maria Rinkl (cropped).
<https://commons.wikimedia.org/wiki/File:Blüte-Gelbe_Teichrose.jpg> License Creative Commons BY-
SA 4.0 International.
A passage in Euripides that is misinterpreted today confirms that capitals represent
projectiles. The chorus in Bacchae (lines 592-593) says, about an earthquake that Dionysus
produces in the house of Pentheus: εἴδετε λάινα κίοσιν ἔµβολα / διάδροµα τάδε; ‘Do you see
these stone missiles on the columns reeling?’. Attending to its etymology (from ἐµβάλλω ‘to
throw into’) the ἔµβολα should not be translated as ‘lintels’, as has been the case, neither as
‘spurs’, which would not make sense, but rather as ‘projectiles’; and we must conclude that
they refer to the stone echini of the columns, not the lintels, which do not have the form of a
Way (1898), like many others, translates: “Lo, how the lintels of stone over yonder pillars are
ram or projectile. Surely because it was the place for the projectiles, the word ἔµβολος ended
up describing a portico attached to a building.
Finally, the archaeologists who excavated the aforementioned Neorion at Delos, called
it the Monument of the Bulls. This was because the triglyphs and the external part of its
capitals had prominent heads of this animal (figure 50). Identifying the capitals with bull's
heads may have been a way of indicating that this naval complement pierced the hull of
enemy ships like the horns of a bull would do.
Figure 50. Capital of the Neorion at Delos, with bull heads in the outer half, detail (left) and lateral view
(right, with cast shadow laterally inverted due to an error of the draftsman). Drawing: Cockerell (1830,
Plate 5, ﬁgs. 3 and 8). Public domain.
The name ἐχῖνος ‘sea urchin’ is due, as mentioned earlier, to the shape of the round
anchor that was holed in the center, recalling the skeleton of this animal; its shape also makes
it similar to an abacus bead, albeit bigger. The presence of an ἄβαξ ‘reckoning-board’ on the
echinus confirms three conclusions reached in previous pages: (a) the echinus, which equates
Liddell, Scott & Jones 1940, s.v. ἔµβολος A.8, and Rodríguez 1989–2019, s.v. ἔµβολον IV.4.
to an abacus bead, is holed in the middle (b) it represents a unit of value or weight, probably
an ingot, and (c) the structure to which the board and echinus belong is inverted, because in
the real world the abacus beads rest on the reckoning-boards, rather than hang from them.
17. The Ionic Column
The Ionic column has hardly any entasis and is thinner and taller than the Doric, making it
probable that the object that inspired it did not support a great weight on top of it. The most
logical thing is that, once the use of the ram made the throwable dolphin obsolete in warships,
the stylos continued to be used, but for holding an emblem and pennant, for identification or
heraldic reasons. We know of the stylis of the stern of a Rhodian galley depicted in a life-
sized bas-relief in Lindos, on the island of Rhodes. It has a small sphere at the top and we
know that Phoenician ships had a sphere finial and a crescent moon as symbol of the goddess
Astarte. The Ionic column probably represents a naval stylis of this type, tall and thin, with
a metallic emblem at one end.
The base of the Ionic column consists of two tori, and a trochilos τροχίλος in the
middle. The term trochilos, in addition to the base’s central moulding, designates the sheave,
the rotating part at the center of a pulley. From Vitruvius (3.5.2) we know that the central
moulding was also called σκοτία ‘darkness’. The central part of a wooden pulley is darker tan
the two lateral parts, either because it is made of iron, or because being made of wood, the
friction of the ropes has burnt its surface. In the case of the Ionic column’s base representing a
pulley, the name of the two tori would come from the adjective τορός ‘holed’, and not from
According to a famous passage of Diodorus Siculus (5.35.4), the Phoenicians obtained so much
silver in their trade with the Iberians that they even substituted the lead anchors of their ships for
others of that precious metal.
Lorenzo 2015, 127.
the noun τόρος ‘drill’, because the parts of the pulley are punctured in the center to allow the
central shaft to pass through them. That the stylos rested on a pulley (or half if the upper
moulding has been eliminated, as appears to be the case, because it is narrower) suggests that
in some circumstances it was necessary to make the stylos and its emblem rotate.
The fluting of the shaft may be the result of reducing the weight of the wooden pole
without losing its resistance to breakage, as happens with the nervation of the stalks of many
plants. The standard number of a shaft’s grooves was 24. The fluting was known in Greek as
ῥάβδωσις, from ῥάβδος ‘wand’. If we imagine a column with 24 rods in the grooves of the
shaft, held by a ribbon, the similarity of this with the fasces that the Romans inherited as a
symbol of authority from the Etruscans will become evident. The fasces were topped with a
double axe, which recalls the Cretan double axe and the axe and panther-shaped finial of the
sceptre found in Malia (figure 40), all of which suggests that effectively the Ionic column
derives from a throwable dolphin and was a symbol of authority.
18. The Ionic Capital
The Ionic capital derives from Proto-Ionic capital, sometimes known as Aeolic, like the one
found at Neandria (figure 51 center). Before tackling its meaning, I would like to point out a
significant difference between one and the other. In the Neandria capital the volutes have
begun to take shape and the palmette is whole, and a little flattened in the center. However, in
the Ionic capital there is a complete flattening: the volutes have experienced a supplementary
rolling, so that they are directed downward and separated inward and outward on both sides of
the entablature line; also the the palmette has been split in half and rolled up together with the
volutes until it is lodged in its axillae (figure 51 right). The logical thing is to attribute the
complete flattening to the effect of the weight of the building on the shape of the capital,
which suggests that the capital—or, rather, the emblem that this one wishes to reproduce—is
made of metal, because only this material has that plasticity lacking in stone and wood. The
front and back of the capital are flat, suggesting that originally it was a flat and possibly thin
emblem similar to the plant-shaped acroteria that crown the pediments of certain temples; the
sides have vertical lines that may be an indication that, in order to create a thick capital,
several finer emblems had to be put together.
Figure 51. Comparison of the top of a Syro-Hittite sacred tree (ﬁgure 53), Proto-Ionic (Neandria) and
Ionic capitals, colored to show the evolution of their parts. Line drawing from the author. License
Creative Commons BY-NC-SA 4.0 International.
One possible objection to the previous argument relating to the flattening is that many
altars depicted in the iconography have volutes that are partial or totally rolled up and
separated palmettes and, however, have never supported a building. In effect, the altar in
figure 52 looks very much like the upper part of an Ionic column, but its deformation, less
than that of an Ionic column, may be due to a different cause. On the altar are deposited, spilt
and burned, offerings that the gods, invisible, will gather up. As we saw previously (Hom.
Hymn VII to Dion. line 18), the Greeks imagined their gods to be bigger and heavier than
humans and the deformation of the altar may be a way of showing that this is a place
frequented by a divinity, which has flattened it with its weight.
See that from the Delphi Museum, Inv. No. 365 (Herdt 2013, 202).
Figure 52. Altar with Ionic volutes, and Athena playing the kithara. Amphora by the Nikoxenos Painter,
c. 500 BC (Staatliche Museen zu Berlin, Preussischer Kulturbesitz F 2161). <https://chs.harvard.edu/
CHS/media/viewImage/1849>. Drawing © Power (2010, plate 11) License Creative Commons 3.0.
Today, we do not know the meaning of the Ionic capital of volutes and palmettes,
which has been documented in the form of an ornament in Mycenaean and Cycladic
objects. It is clear that it is a variant of the Assyrian symbol of the Sacred Tree, whose
meaning in turn scholars have been discussing for a century and a half without reaching an
agreement. Next I will propose a totally new interpretation of this symbol based on a Syro-
Hittite relief from the 11th century BC. found on the hill at Arslantepe (ancient city of
Wilson Jones 2014, 96-111.
The identification is due to Wurz (1914, 64), and confirmed by Akurgal (2001, fig. 151b).
Regarding the so-called ‘Proto-Aeolian’ capitals found in Palestine and Cyprus, the same origin
is generally accepted (Shiloh 1977, 46).
Manuelli and Mori 2016, 226.
Melid) on the Upper Euphrates (figure 53). This sacred tree is flanked by two eagle-head
hybrid-demons armed with a club and a curved object. The tree may be a reference to the
weapons used in war. The rods wielded by the demons, are usually interpreted as
lightnings, and are similar to those of the top of the tree. It is consistent with this
interpretation to assume that the rods represent the wooden shaft of more ordinary throwing
weapons, such as spears or arrows which, after all, are the poor version of thunderbolts fired
by the gods. In Hittite pašk-/paški(ya)- means ‘branch’ and also ‘arrow’, and is specifically
used to refer to the branches of the eya tree, the tree of life or sacred tree. In other
contemporary sacred trees, like in many Ionic capitals, the upper part consists rather of radii
that start from a circle that is partially visible between the volutes. In my opinión, this motif
represents, not a palmette, but a sun disk or a flower of the Asteraceae or Compositae type,
such as the field marigold (Calendula arvensis). In fact, in Greek the word δίσκος has, among
others, these three meanings: (a) throwable quoit, (b) solar disc and (c) field marigold. If
we stick with the interpretation of the plant element as a weapon, we must conclude that the
flower (or the sun o the star, which these interpretations also fit) represents a projectile.
Manuelli and Mori 2016, 227 and note 28.
Güterbock and Hoffner 1997, s.v. pask-: “They plant twigs of the eya-tree”.
Liddell, Scott & Jones 1940, s.v. δίσκος. The association of the sun’s rays with rods is ancient and
obvious, and in Latin radius means both ‘rod’ and ‘ray of sun’.
Figure 53. Relief of two eagle-head hybrid-demons, Arslantepe (9th century BC.). Author: Zeynel
License Creative Commons BY-SA 4.0 International.
The volutes below the rods in the Arslantepe sacred tree can represent the curved finial
of the two limbs of a composite bow. Like the shafts of arrows and spears, the limbs of a bow
are also made from tree branches, but not from straight branches, but from bent ones.
Below the volutes there is a conical piece that surely represents the trunk of what ultimately is
depicted as a tree. Lastly, the overlapping triangles of the lower part may represent the
sharpened metal blades of different types of weapon: in order from largest to smallest,
swords, daggers, spearheads and arrowheads. The capitals found in Palestina that have been
called Proto-Aeolics have these triangular forms below the volutes. The straight and bent
rods are presented as branches of the sacred tree, but the metal blades do not come from any
tree, but are obtained from minerals extracted from the earth. That is why they are shown
below the trunk, forming a mineral base to which the trunk clings, in the same way as a tree
grows roots in the earth. The sacred tree would be, then, a military symbol representing
weapons and—using the part for the whole—the army that is at the king’s disposal.
Other sacred trees, as one from the Northwest Palace of Ashurnasirpal II (Layard 1849, plate 7),
show completely symmetrical branches finished with volutes of an upwards arrangement, more
similar to composite bows. In the same tree, there are some horsehair helmet crests. In other
sacred trees (plate 6), there are poles topped with what look like pines or fruits, which may
represent clubs. Several sacred trees are flanked by genies and kings who, using a bucket and a
cone, fertilize them manually (plate 25). It may well be a metaphor, representing how the effort
of kings and the favor of the divinity increase the number of soldiers and weapons in the army.
Seidl and Sallaberger (2005–2006, 54–74) maintain that the sacred tree is the object which
Assyrian written sources refer to as urigallu. This was a cult object, made sometimes of reed,
used in some rituals as a protector, of the gates and of a sick man among others (Wiggermann
1992, 70–73). The term urigallu is a Sumerian loan, language in wich it literally means ‘great
protector’, from ùri ‘protector’ and gal ‘great’ (Wiggermann 1992, 70). In Assyrian it had the
meaning of ‘emblem, standard’, and actually many Palestinian and Greek volute capitals are at
the top of orthostats or non-architectural columns. All this endorses the interpretation that we
propose here of the Ionic column as a naval standard or stylos.
Shiloh 1979, passim. Many Proto-Aeolic capitals are on a pair of columns flanking an entrance
door, which allows them to be identified as a representation of two armed soldiers or geniuses
guarding the door (see previous note). Earlier, Assyrian texts mention the urigallu (‘standards’)
in this position: ‘they set up the guardians of the gate, the standards, to the right and left of the
gates’ (Thureau-Dangin 1921, 120 line 26; trans. Roth 2010).
As for the Proto-Ionic and Ionic capitals, the palmette and the volutes most probably
represent the compound bow and the arrows used by the ten archers embarked on the Ionian
ships of the Archaic period (Hdt. 3.39.3). But as this symbol was sculpted in Archaic Ionia in
a religious context, it may also represent the arrows and bow of the twin gods Apollo and
Artemis, to whom several famous temples were dedicated in the region, such as that of Apollo
at Didyma and of Artemis at Ephesus. Both explanations are compatible.
Below the volutes, in many capitals we find the well-known moulding of egg-and-
dart. At first sight, the eggs could represent slingshot missiles and the darts arrowheads and/or
spearheads, but a closer look at capitals such that of the votive column at Aphaia reveals
that the piece is a flower-shaped echinus, thus inherited from the Doric capital, and that the
‘eggs and darts’ are the ends of petals and sepals of the flower.
From all the above it follows that the Ionic capital constitutes the peculiar Greek
version of the Near-Eastern Sacred Tree.
19. The Corinthian Capital
The Corinthian Capital arises at the beginning of the 4th century BC, at the same time that a
revolution in the art of war was taking place, a revolution that would change the development
of sea warfare for centuries. According to Biton of Pergamon (Construction… 61), during the
siege of Mileto in 421 BC the γαστραφέτης ‘belly bow’ was used for the first time. This was a
composite bow that was loaded by first pushing it with the belly against the ground and later
It could be as well that the volutes of the Ionic capital had, before the total flattening of the capital,
the shape of a crescent moon, and the upper floral motif represented the sun. In this case, it
would be a composed symbol of the gods Apollo and Artemis, or the same symbol (sphere and
crescent) that crowned the stylos of the Phoenician ships which, as we have already mentioned,
is related to the goddess Astarte (Lorenzo 2015, 127).
See Herdt 2013, 207.
with a crank. It shot large arrows, but later Charon of Magnesia built version known as
πετροβόλος or λιθοβόλος that fired stones of 2 to 3 kg in weight (Biton 45-47) (figure 54).
The perfecting and increase in size of this weapon led to the invention of the catapult in Sicily
in 399 BC (Diod. Sic. 14.42.1), which is usually considered to be the beginning of artillery.
The catapult, first used in sieges of cities, was later installed on ships and became the most
powerful weapon for distance fighting in naval battles.
Details on the construction of the petrobolos are discussed in Campbell and Delf (2003, 5–7).
Further up in this text we have proposed the shaduf as the ultimate origin of the naval dolphin, a
simple device of gravity artillery. Coincidentally, there is agreement among scholars that the
shaduf is the origin of the trebuchet, the traction and counterweight catapult; however, this siege
machine was invented in China in the 5th century BC, and in the West it did not begin to be used
by the Byzantines until the 6th century AD (Petersen 2013, 406-408), according to the testimony
of Maurice in Strategikon 10.3.
Marsden (1969, 169-173). We know for certain that the catapult was used on board a ship around
332 BC, when Alexander besieged Tyre (Diod. Sic. 17.42.1-2), but his is only the first time such
a use is mentioned however, and seaborne catapults may have been used earlier. For the length
and weight of projectiles, see Murray (2012, table 5.1).
Figure 54. Petrobolos, the stone thrower of Charon of Magnesia. Line draw by the author, based on a
scale model by Kotsanas, n.d. License Creative Commons BY-SA 4.0 International.
The Corinthian capital, if it replaced the emblem of volutes and palmettes at the top of
naval styloi, it did so because of the meaning that its components transmitted. Acanthus
spinosus is a plant that stands out in Mediterranean flora for the way in which it disperses its
seeds, through a process known in botany as explosive autochory, ballochory or ballistochory,
where “the stalks of the ovules (…) act as a catapult, throwing out the seeds”. The launch
occurs thanks to the action of the devices called iaculatores ‘throwers’ (figure 55), which
throw the seeds up to 10 meters from the plant.
Polunin and Everard 1991, p. 393.
B & T Worlds Seeds, n. d.
Figure 55. Acanthus mollis seeds in their capsule and, at right, a iaculator after having shot the seed.
Drawing by Engler (1895, ﬁg. 127 M and N). Public domain.
As for the volutes of the Corinthian capital, their shape recalls the tendrils, the stalk
extensions, of some creeping plants. It is not important to identify the species the tendrils
belong to, because their role in the capital is to represent elasticity in general, stretching,
releasing and then immediately returning to their original form. The tendrils would represent,
therefore, the horns and arms of the gastraphetes and the first catapults, such as the
petrobolos, which build up energy through the tension of the limbs of a composite bow.
According to Vitruvius’ well-known story (4.1.9-10) in which the invention of the
Corinthian capital is attributed to Callimachus, in the center of the capital there was originally
a κάλαθος ‘basket’, which gave the capital its shape of a tube that is narrower at the bottom.
The shape of the capital may in fact represent a basket in which stone projectiles for the
petrobolos were kept.
At the top of each of the Corinthian capital’s faces there is a flower that in my opinion,
very much resembles the blossom of the round gourd (κολοκύνθη, Cucurbita maxima), and to
a lesser extent of the colocynth (κολοκύνθη ἀγρία, Citrullus Colocynthis L.), and would
represent the perfectly spherical stones hurled by the petrobolos. In the center of the flower in
the Corinthian capitals of the Temple of Olympian Zeus at Athens, circa 178 BC, there is what
appears to be a canteen, with lid included, like those made from the fruit of the squash. The
capitals of the Roman Porticus Octaviae, circa 200 BC, instead of a flower have an eagle
holding a thunderbolt in its claws, confirming that this element of the capital represents a
projectile. In other later Roman capitals, such as those of the Temple of Vespasian and
Titus, the flower has at its center a sinuous figure that may represent the body and tail of a
fish that is collecting pollen (figure 56). Since it is impossible for it to be a realistic
representation, then it may instead be an ingenious way of identifying the center of the flower,
i.e. the projectile launched by the catapult; it is so hard that an iron dolphin that crashes into it
See Penrose 1888, pl. 39. The ovary of the yellow water-lily (Nuphar lutea), also known as
‘brandy-bottle’, is bottle-shaped, but has nothing similar to a lid. Its petals are different from
those represented in that capital.
See Desgodetz 1682, p. 171, who wrongly attributes the portico to Septimius Severus.
Similar flower decoration in the capitals of the Temple of Antoninus and Faustina and the Pantheon
Figure 56. Flower of an unknown Roman corinthian capital with what may be a dolphin collecting
pollen. Drawing: anonymous author, 16th century. <https://commons.wikimedia.org/wiki/File:Unknown,
Corinthian capital, elevation (recto) Unknown, Corinthian capital, plan diagram and detail (verso) MET
DP810838.jpg> License Creative Commons CC0 1.0 Universal.
The first known Corinthian capital is that of the temple of Bassae, built between 450
and 400 BC, contemporary with the invention of the gastraphetes. It is the only capital in the
temple, and the column it crowns is in a very prominent position, in the center of the cella, in
front of the statue of Apollo if seen from the door. The column covers it and occupies the
central place that would correspond to it (figure 57), and as a result the temple has an odd
number of columns, contrary to the usual practice. If the statue of the god represents a large
weight used as a projectile in the naval battles using only the force of gravity, its partial
concealment by the column with Corinthian capital may signify the emergence of a new form
of artillery at sea: the catapult.
Figure 57. Plan of the temple of Apollo at Bassae, with Doric (turquoise), Ionic (olive) and Corinthian
columns (yellow). Line drawing by the author. License Creative Commons BY-NC: SA 4.0
Most temples have all their columns and capitals of the same style, but in the Bassae
temple there is a particular combination of Doric capitals (in the peristasis, in the pronaos and
the opisthodomos), Ionic (in the interior of the cella) and the new Corinthian capital. In a
previous section we have seen how the columns represented the crew of the penteconter and
that their location identified them as rowers (those of the peristasis) or sailors and soldiers
(those of the cella). The combination of Doric, Ionic and Corinthian styles makes sense in
light of this interpretation. The old stylos with a stone anchor at the top that was used to break
through the hulls of enemy ships was replaced by the ram; the rowers, officers and sailors
who move the ship and push it against the enemy ship’s hull fight, so to speak, with the ram,
and are represented by columns and Doric capitals arranged mainly in the perimeter of the
overturned ship. The ten soldiers in the interior of the ship fight with their bows and are
represented by Ionic columns and capitals. In this temple, for the first time, there is a single
As we saw earlier, there were archers in the fleet of Samos in the 6th century BC (Hdt. 3.39.3),
even though in the Athens of the 5th century BC, they had been replaced by hoplites (Thuc.
additional central column which represents the soldier who fires the petrobolos, or the
20. The acroteria of the pediments
The pediment of the temple’s western side, opposite the entrance usually found on the eastern
side, is what remains after having sectioned the overturned ship’s bow, the place where the
ἴκρια ‘forecastle’ was. But upon sectioning not only the stern but also the bow of the
overturned ship, the two short ends of the temple ended up being equal. Now, thanks to our
interpretation of the capitals as representations of the different methods of naval combat, we
can now tackle the meaning of the acroteria that crown the pediments. On the central and
lateral platforms of certain temples there are acroteria that adopt any of the following forms:
(a) a large disc with concentric lines, such as the temple of Aphaea at Aegina, (b) acroteria of
volutes and palmettes, (c) sphinxes or (d) anthropomorphic figures. According to what we
have previously concluded, three of these (the disc, sphinxes and anthropomorphic figures)
represent anchors or ingots, that is, great weights, which were used as a throwable dolphin,
while the acroteria of volutes and palmettes represented the use of the bow.
21. Bow statues as anchors
Up to this point we have identified the use of large weights in the shape of an animal or
anthropomorphic god as a projectile, and to a lesser extent as an anchor. We know from its
name (ἱερὰ ἄγκυρα ‘sacred anchor’) that the anchor reserved for extreme situations was
associated with the gods as a way of ensuring their protection, and that although it would be
exceptional, this association would be made explicit by giving the anchor the form of an
anthropomorphic god. We confirmed this earlier by analyzing the Homeric Hymn VII to
Dionysus, in which a sculpture of Dionysus, really a naval projectile, is confused with an
anchor. In Archaic and Classical Greece, the preferred metal would have been bronze cast
with the lost-wax method. Most preserved bronze statues have been found on the sea floor,
and we assume that they all formed part of the cargo of shipwrecked ships. The famous
Winged Victory of Samothrace, however, may represent a bronze sculpture carried in the
prow as a sacred anchor. The Croatian Apoxyomenos was found on the sea floor with no
indications of shipwreck, so it could have been used as an anchor tied to a provisional cable
and lost. Something similar may have happened to the Riace Warriors.
In the temple, the statue is situated, as in the ship, on the axis of longitudinal
symmetry, and on the bow, that is, on the furthest side to the entrance, which coincides with
the ship’s stern. In addition to its divine character, its size and construction with heavy
materials leave no room for error: they are anchors for survival at sea and projectiles for
victory in war. In front of the chryselephantine statue of Athena in the Parthenon there was a
small, shallow pool, full of water. The statue of Olympian Zeus had an identical one full of
oil. Pausanias believes (5.11.10) that the pools served to maintain the necessary humidity to
keep the ivory of the statues in good condition. Some present-day scholars, however, reckon
that the purpose was aesthetic: the light that entered through the door would be reflected in
the surface of the water or oil and would produce an unusual illumination of the statues.
Both interpretations may be true and are compatible, but there is room for another
So far we have barely mentioned the mainmast, which is a prominent element of a ship. Possibly
the architects decided not to represent the mast in any way, because before the combat it was
picked up, and from that moment on warships were propelled only by oars. However, it is
possible that the large cult statues such as Athena in the Parthenon and Zeus in Olympia
represent it in some way, since they were built by attaching pieces to a kind of central mast.
Miles 2016, 213.
explanation: these statues, if they were sacred anchors like the Winged Victory of Samothrace,
they would have been looked at originally on the prow of a ship, and their image reflected in
the surface of the water. The pools of the two temples served, therefore, to reproduce the
original experience, seeing the image and its reflection, and also could have been to achieve
the effect of sympathetic magic: while the surface of the pool of water or oil was still, the
surface of the sea would also be still, contributing to a safe voyage for the sailors. The choice
of oil in Olympia is surely due to the fact that it is more viscous tan water, so that its surface
would have stayed as still “as—to use a common cliché—a pool of oil”.
22. Expanding the visual field
From what has been said so far we can conclude that Greek temples are more than buildings:
they are stone sculptures of ships which unequivocally proclaim the Greeks’ familiarity with
the sea and their mastery of it, their thalassocracy. So far we have referred almost
exclusively to the Greek temple created between the Geometric and Classical periods, but this
is only part of a much longer history. At the beginning of this article we discussed indications
that, during the second and first millennia BC, at least the inhabitants of the Balearic Islands,
some of the Sea People and the inhabitants of Lycia had or knew of the custom of setting up
camp underneath overturned ships, a custom that was probably widespread in the
Mediterranean. It is also likely that the use of overturned ships as a ceiling, as well as the use
of anchors as missiles, can be traced back centuries to the Egyptian culture, but that oversteps
the scope of this study.
It is surely no coincidence that we refer to the combined legacy of Greeks and Romans as classica
cultura, meaning in Latin ‘the culture of the fleet’.
However, it is possible to produce a much wider sketch than we have done so far of
the history of the presence of overturned ships in Aegean culture itself. Troy II-III, around
2200 BC, is one of the earliest examples of the use of the plan that we know as µέγαρον
(figure 58). Until now we have not mentioned these buildings that for decades were
considered by the scientific consensus to be the Mycenaean precursors of the Greek temple,
and we have focussed on a creation ex novo of the Greek temple during the Geometric period.
It is likely that similar developments took place at two different moments in history: of
overturned ships evolving into palaces in the Mycenaean period and, after the hiatus of the
Dark Age, of overturned ships evolving into temples in the Geometric period. Later we will
see that, surprisingly, the same process was repeated for a third and fourth time in other times
Figure 58. Plan of the citadel of Troy II. Drawing courtesy of © Elizabeth Riorden (Rose 2013, ﬁg. 1.4).
The plans of the central buildings at the citadel of Troy II-III are very similar to the
plans of the elongated buildings of the Geometric period: the two structures that flank the
principal megaron may have supported ships that could fit edgewise through the door of the
entrance to the citadel; not so the central building. The fortified citadel was situated at that
time by the sea, and the southeastern door (fig. 58, FM) opened to a ramp that led to the
beach. This could have enabled them to pull the ships of the king into the interior of the fort in
order to store them during the winter. It is no surprise that the ships were stored there, since
they represented, along with the walls, one of the biggest investments in military equipment
of the age. In Classical Athens, a procession recalls, in my opinion, the act of bringing up
the city’s flagship to the Acropolis and storing it there. During the quadrennial procession of
the Great Panathenaea, a peplos decorated with the victory of Athens in the Gigantomachy,
the size of a sail, was hung from the yardarm of the Panathenaic Ship. This ship was carried in
procession through the streets of Athens before finally being taken up to the Acropolis.
We have focussed on explaining the naval origin of the temple, but the cities of the
Geometric period had dozens of ships, and others ended up giving rise to similar public
buildings: στοαί ‘stoas’, which would have been created with a single ship or by placing
several ships in a row, ἑστιατήρια ‘public eating places’, the σκενή ‘tent, scene’ of the theater
which, as we know, represents a palace from heroic times, etc. The term στοά has the typical
acute accent of an adjective, and is formed from the root ἵστηµι ‘stand up’; it requires,
therefore, a feminine noun, which would plausibly be ἡ ναῦς ‘ship’. The expression ἡ στοὰ
ναῦς would then mean ‘the standing ship’, that is, ‘raised on supports’.
Many ancient authors and modern scholars have claimed that the Trojan horse, which the Trojans
brought into their citadel by breaking down part of its walls, was actually a ship; according to
some, of the Phoenician ship type known as hippos (Ruiz de Arbulo 2009, 240–241).
It is likely that the mariners who sailed in flotilla, when setting up camp in hostile
territory, did so by arranging the ships in a closed formation, like a walled perimeter. The
παλαίστρα ‘palaestra’ would be derived from this arrangement and later the ἀγορή, the square
or rectangular ‘agora’ of Hellenic cities and the Roman forum. The ship that served as an
entrance to this enclosure would give rise to προπύλαια ‘propylaea’.
In Athens around 500 BC we see the erection in the agora of ἡ βασιλικὴ στοά ‘the
royal stoa’, which is in the origin of the Latin term basilica; if the derivation proposed above
is true, this building was constructed by placing the royal ship or the city’s flagship on
supports. The Roman basilica however does not derive directly from this building. During the
classical and Hellenic periods, shipsheds for the fleet were constructed in Athens (Zea
Harbour), Carthage and many other cities. These followed a very simple pattern: they
consisted of a series of extended rectangular passageways separated by columns and with
access to the sea. Ships would be raised from the sea and slid into place in the shipshed.
Logically, the flagship’s shipshed was the largest since it had to accommodate what was the
largest ship in the fleet. The design of the Roman basilica derives instead from the simple
shipshed that no longer had any ship on the roof, but in the interior.
But what is the reason for this obsession of the ancient Greeks for living under ships?
Thucydides begins his history of the known world by making a list of the thalassocracies that
arose in Greece before his time, because he understood that that domination of the sea meant
the domination of maritime commerce, the main source of wealth to which the Greeks could
aspire to then as now. If the construction and exploitation of warships and merchant ships was
the source of power and legitimacy of the Mycenaean and Greek elites, it is logical that this
It follows that the first agorae with stoas would be by the port. This is the case of ‘the place of
assembly [agorē] of the Phaeacians, which was builded for them hard by their ships’ (Od. 8.5).
object, the warship, and war at sea, permeated not only artistic manifestations as architecture,
but also other notable aspects of Greek culture such as athletic games and the theater which
we have no space to deal with here. Heraclitus referred to this process with the words ‘War is
the father of all’ (fr. 53 DK).
To confirm our hypothesis it would be definitive to find an ancient overturned ship
that was used as shelter or temple. Well, it has been found: less conspicuously in the stone
tombs of the Lycian coast (figure 10), in the vicinity of Greece, but much more clearly
thousands of kilometres from the original Greek homeland. Hellenism spread Greek culture
far from Greek soil. In the India of the 3rd century BC the descendents of Macedonian and
Greek colonists who lived under the emperor Ashoka the Great converted to Buddhism. As I
have explained in a recent article, they originated a Buddhist architecture repeating for the
third time the process of conversion of overturned ships into buildings. The entrance to the
cave of Lomas Rishi (figure 59) was sculpted around 250 BC in the shape of the hull of an
overturned ship. In my opinion it is the work of Buddhist Indo-Greeks who identified
themselves with Odysseus and his companions in the episode of the arrival at the island of
Helios, the easternmost land they reached, where they say: ‘we dragged our ship, and made
her fast in a hollow cave / where were the fair dancing-floors and seats of the nymphs’ (Od.
12.316-318) . During the next centuries faithful Buddhists excavated cave temples in the
Maharashtra region reproducing the hulls of almost complete overturned ships (figure 60).
Over time the use of the basilical plant and the piling up of ships—this inspired by the
memory of the pyre that Alexander the Great had built in honour of his general Hephaestion
Murray, trans., 1919.
(Diod. Sic. 17.115.2-4)—moulded religious architecture in Asia in the last two thousand
Figure 59. Entry to the Lomas Rishi cave in Barabar (Bihar, India). Photo: Neilshatian. <http://
commons.wikimedia.org/wiki/File: Entrance_Gate.JPG>. License Creative Commons BY-SA 3.0.
Figure 60. Inside of the cave 26 in Ajanta (Maharashtra, India). Photo: Dey Sandip. <http://
commons.wikimedia.org/wiki/File:Cave_26,_Ajanta.jpg>. License Creative Commons BY-SA 3.0.
Something similar, and for the fourth time, occurred in the west. Christianity is a
Hellenistic religion that, as regards architecture and if studied in light of this new
interpretation of Greek architecture, turns out to be much more continuist than we had
previously thought. To give just one advance, the name of the first Christian construction, the
catacumba, is usually said to derive from the Greek κατά ‘downwards’ and the latin tumba,
but this is not a convincing etymology, because it does not explain the supposed phonetic
evolution -t- > -c-; or it is said that it derives from the verb cubo (in composition *-cumbo) ‘to
lie down’. It is much more likely however that the word derives from the Greek κύµβη > lat.
cumba (rarely cymba), a ship of Phoenician origin, which according to Roman writers was the
type of ship used by Charon to transport the dead to the beyond. The term catacumba
would mean, therefore, ‘down the boat’, a meaning that only now, after unravelling the naval
origin of Greek architecture, would be semantically acceptable to us. It would not always
Verg. Aen. 6.303, Hor. Carm. 2.3.28 and others.
have been convenient to raise an overturned ship on supports because it could have fallen. It
could have been left overturned on the ground, and to be able to take shelter in, it would have
been enough to dig underneath it and fashion some steps to access the recently created
hypogeum space. But that is another long story that I would like to tell in a future article.
As regards not the history but the general theory of architecture, the relationship
between form and function in Greek architecture should be revised in the light of this new
interpretation, as well as the consideration, due to Adolf Loos , of the so-called ornament as
[Place], 31 August 2020.
Abbas, E. 2013. “The Tale of the Shipwrecked Sailor and the Rite of Passage”, Journal of the
American Research Center in Egypt 49, 9–18. DOI: 10.5913/0065-9991-49-1-9.
Akurgal, E. 2001. The Hattian and Hittite Civilizations, Ankara: Republic of Turkey Ministry
Alexander, C. 1998. The Endurance: Shackleton’s legendary Antarctic expedition, London.
Anonymous, trans. 1928. Aristophanes: The Eleven Comedies, New York: Horace Liveright.
Anonymous, Naval Warfare, commissioned by Basil, the patrikios and parakoimomenos, in
The Age of the ΔΡΟΜΩΝ. The Byzantine Navy ca 500–1204, J. H. Pryor and E. M.
Jeffreys, eds., Leiden and Boston, App. 3, 521–545.
B & T Worlds Seeds, n. d. “Acanthus mollis. Bear’s Breaches Acanthaceae”. Retrieved from
https://b-and-t-world-seeds.com/13.htm [Last accessed 18 December 2021].
Bankel, H. 1993. Der spätarchaische Tempel der Aphaia auf Aegina, Berlin and New York.
Basch, L. 1987. Le musée imaginaire de la marine antique, Athens.
Beekes, R. 2010. Etymological Dictionary of Greek, Leiden and Boston.
Bostock, J. and H. T. Riley, trans. 1857. The Natural History of Pliny, vol. VI, Princeton: H.
Boutsikas, E. 2009. “Placing Greek Temples: An Archaeoastronomical Study of the
Orientation of Ancient Greek Religious Structures”, Archaeoastronomy 21, 4–19.
Bowen, M. L. 1950. “Some Observations on the Origin of Triglyphs”. The Annual of the
British School at Athens 45, 113-125. DOI: 10.1017/S0068245400006729
Campbell, D. B. and B. Delf. 2003. Greek and Roman Artillery 399 BC–AD 363, Oxford.
Cartwright, M. 2012. “Hercules”. World History Encyclopedia. 9 July 2012. Retrieved from
https://www.worldhistory.org/hercules/ [Last accessed 18 December 2021].
Casson, L. 1995. Ships and Seamanship in the Ancient World, Baltimore and London.
Chantraine, Pierre. 1968. Dictionnaire étymologique de la langue grecque, Paris.
Ciordia, J. M. 2005a. “The Naval Origin of Greek Architecture and Sculpture: Plan of the
Book” (available online https://pompilos.org/file_download/1/, accessed 16-08-2005).
———. 2005b. “El origen naval de la arquitectura y la escultura griegas. Capítulos 1, 2 y 3”
(available online https://pompilos.org/file_download/3/, accessed 16-08-2005).
———. 2007. “Del shadouf a las columnas dórica y jónica” (available online https://
pompilos.org/file_download/4/, accessed 25-10-2007).
———. 2020. “The Ship in the Cave: The Greek and Nautical Origin of Buddhist
Architecture”, Journal of Asian Architecture and Building Engineering, 19:1, 48–69.
Coates, J. F. 1995. “The Naval Architecture and Oar Systems of Ancient Galleys”, in The Age
of the Galley. Mediterranean Oared Vessels since Pre-Classical Times, R. Gardiner
and J. Morrison, eds., London, 127–141.
Coates, J. F. 1999. “Long Ships, Slipways and Beaches”, in Tropis V. 5th 7th International
Symposion on Ship Construction in Antiquity. Nauplia 1993. Proceedings, H. Tzalas,
ed., Athens, 103–18.
Cockerell, C. R., W. Kinnard, T. L. Donaldson, W. Jenkins and W. Railton. 1830. The
Antiquities of Athens and other places in Greece, Sicily etc., supplementary to the
Antiquities of Athens by James Stuart F.R.S. and F.S.A. and Nicholas Revett, London.
Coleman, J. E. 1985. ‘“Frying pans” on the Early Bronze Age Aegean’. American Journal of
Archaeology 89:2, 191–219. DOI: 10.2307/504325.
Coulton, J. J. 1977. Greek Architects at Work: Problems of Structure and Design, London.
Covarrubias, Sebastián de. 1611. Tesoro de la lengua castellana o española, Madrid: Luis
Crawley, R., trans. 1910. Thucydides: History of the Peloponnesian War, London and New
Del Freo, M. 2005. “L’expression ka-ko na-wi-jo de la tablette Jn 829 de Pylos”, in Emporia :
Aegeans in the central and eastern Mediterranean: proceedings of the 10th
international Aegean Conference = 10e rencontre égéenne internationale, Athens,
Italian School of Archaeology, 14-18 April 2004, R. Laffineur and E. Greco, eds.,
Liège and Austin, 793–803.
Desgodetz, A. 1682. Les édifices antiques de Rome, Paris.
Dickens, C. 1850. David Copperfield, with illustrations by H. K. Browne, Philadelphia.
Dietrich, M. and O. Loretz. 1978. “Das “seefahrende Volk” von Šikila (RS 34.129)”, Ugarit
Forschungen 10, 53–56.
Edwards, A. B. 1890. A Thousand Miles up the Nile. London.
Engler, A. 1895. Die Natürlichen Pflanzenfamilien 4 (3b), Leipzig.
Evelyn-White, H. G., trans. 1914. Hesiod, the Homeric hymns and Homerica, London and
Fair, S. W. 2005. “The Northern Umiak: Shelter, Boundary, Identity”, Perspectives in
Vernacular Architecture 10, 233–248. https://www.jstor.org/stable/3514351
Frost, H. 1991. “Anchors Sacred and Profane. Ugarit-Ras Shamra, 1986; the stone anchors
revised and compared”, in Ras Shamra–Ougarit VI: Arts et industries de la pierre, M.
Yon. (ed.), Paris, 355–410.
Gebhard, E. R. 2001. “The Archaic Temple at Isthmia: Techniques of Construction”, in
Archaische Griechische Tempel und Altägypten, M. Bietak (ed.), Wien, 41–61.
Gianfrotta, P. A. 1997. “La piraterie”, in La navigation dans l’Antiquité, P. A Gianfrotta, X.
Nieto, P. Pomey and A. Tchernia, eds., Aix en Provence, 46–57.
Giovino, M. 2007. The Assyrian Sacred Tree: A History of Interpretations, Fribourg and
Guerrero, V. M., M. Calvo and S. Gornés. 2006. El poblamiento prehistórico de las Islas
Baleares, Palma de Mallorca.
Güterbock, H. G. and H. A. Hoffner, eds. 1997. The Hittite dictionary of the Oriental Institute
of the University of Chicago, Volume P, Chicago.
Harmon, A. M., trans. 1913. Lucian vol. I, Cambridge, Mass.
Herdt, G. 2013. Votive Columns in Greek Sanctuaries of the Archaic Period (Unpublished
doctoral dissertation), Bath: University of Bath.
Höckmann, O. 2002. “Stern Rams in Antiquity?”, in Tropis VII. 7th International Symposion
on Ship Construction in Antiquity. Pylos 1999. Proceedings, I, H. Tzalas, ed., Athens,
Hurley, F. 1910-1962. “View of interior of hut on Elephant Island (Shackleton expedition,
1916)” [Picture]. National Library of Australia. Retrieved from https://nla.gov.au/
nla.obj-160222368, accessed 26-04-2022.
Julistiono, E. K. and L. S. Arifin. 2005. “The sustainable traditional structural system of
‘Tongkonan’ in Celebes, Indonesia”, in SB05 Tokyo: Action for Sustainability - The
2005 World Sustainable Building Conference, Tokyo, 27–29 September 2005,
Kapitän, G. 1984. “Ancient Anchors—technology and classification”, The International
Journal of Nautical Archaeology and Underwater Exploration 13.1, 33–44.
King, C. 1970. “The Homeric Corslet”, American Journal of Archaeology 74:3, 294–296.
Kirk, G. S. and J. E. Raven. 1971. The Presocratic Philosophers, London and New York.
Komber, J. 2001. “Viking Age Architecture in Space and Time”, Ruralia 4, 13–29.
Kotsanas, K. n.d. “The stone-throwing catapult of Charon of Magnesia” (available online
http://kotsanas.com/gb/exh.php?exhibit=1402003 accessed 18-08-2020).
Kyrieleis, H. 1993. “The Heraion at Samos”, in Greek Sanctuaries: New Approaches, N.
Marinatos and R. Hägg, eds., London and New York, 125–153.
Lavalou, A. 1998. “Retour aux sources (Back to origins)”, L’architecture d’aujourd’hui 317,
Layard, A. H. 1849. The Monuments of Nineveh I, London.
Liddell, H. G., R. Scott and H. S. Jones. 1940. A Greek-English Lexikon, Oxford.
Loos, A. 1913. “Ornement et crime”, M. Ray trans., Les cahiers d’aujourd’hui 5 (June), 247–
Lorenzo, K. 2015. “Triremes on land: First-fruits for the Battle of Salamis”, in Autopsy in
Athens: Recent Archaeological Research on Athens and Attica, M. M. Miles, ed.,
Oxford and Philadelphia, 126–138.
Manuelli, F. and L. Mori. 2016. “‘The King at the Gate’. Monumental fortifications and the
rise of local elites at Arslantepe at the end of the 2nd millennium BCE”. Origini 39,
Marchant, J. 2016. Human skeleton found on famed Antikythera shipwreck. Nature 537, 462–
463. DOI: 10.1038/537462a
Mark, S. 2008. “The Earliest Naval Ram”, International Journal of Nautical Archaeology
37:2, 253–272. DOI: 10.1111/j.1095-9270.2008.00182.x
Marsden, E. W. 1969. Greek and Roman Artillery. Historical Development. Oxford.
Mazarakis Ainian, A. 1997. From Rulers’ Dwelings to Temples. Architecture, Religion and
Society in Early Iron Age Greece (1100–700 B.C.), Jonsered.
McEwen, I. K. 1993. Socrate’s Ancestor. An essay on Architectural Beginnings, Cambridge,
Mass. and London.
Miles, M. M. 2015. “The Vanishing Double Stoa at Thorikos and its Afterlives”, in Autopsy in
Athens: Recent Archaeological Research on Athens and Attica, M. M. Miles, ed.,
Oxford and Philadelphia, 163–180.
———. 2016. “The Interior of the Greek Temples”, in A Companion to Greek Architecture,
M. M. Miles, ed., Chichester, West Sussex, 206–222.
Morgan, M. H., trans. 1914. Vitruvius: The Ten Books on Architecture, Cambridge, Mass.
Morrison, J. 1995. “The Trireme” in The Age of the Galley. Mediterranean Oared Vessels
since Pre-Classical Times, R. Gardiner and J. Morrison, eds., London, 49–65.
Murray, A. T., trans. 1919. Homer: Odyssey, Cambridge, Mass.
Murray, W. 2012. The Age of Titans. The Rise and Fall of the Great Hellenistic Navies. New
Onians, J. 2002. “Greek Temple and Greek Brain”, in Body and Building: Essays on the
Changing Relation of Body and Architecture, G. Dodds and R. Tavernor (eds.),
Cambridge, Mass., 44–63.
Pantazis, V. D. 2013. “Από τη ‘νῆα’ στο ‘ναό’”, The Books’ Journal 38, 92–95.
Pearson, C. L., P. W. Brewer and others. 2018. “Annual radiocarbon record indicates 16th
century BCE date for the Thera eruption”, Science Advances 4, eaar8241. DOI:
Penrose, F. C. 1888. An Investigation of the Principles of Athenian Architecture, London and
Pernigotti, A. P. 2021. “Etruscan Temples and the Sun: An Analysis on the Orientation of
Etruscan Sacred Buildings”, in Beyond Paradigms in Cultural Astronomy:
Proceedings of the 27th SEAC Conference Held Together with the EAA, A. C.
González-García and others (eds.), Oxford, 125–132.
Petersen, L. I. R. 2013. Siege Warfare and Military Organization in the Successor States
(400–800 AD). Leiden and Boston.
Petrakos, B. 1997  “Θορικός”, Το Έργον της εν Αθήναις Αρχαιολογικής
Polunin, O. and B. Everard. 1969. Flowers of Europe: A Field Guide, Oxford.
Pomey, P. 1997. “Les navires”, in La navigation dans l’Antiquité, P. A Gianfrotta, X. Nieto, P.
Pomey and A. Tchernia, eds., Aix en Provence, 60–102.
Power, T. 2010. The Culture of Kitharôidia, Washington D.C.
Pryor, J. H. and Jeffreys, E. M. 2006. The Age of the ΔΡΟΜΩΝ. The Byzantine Navy ca 500–
1204, Leiden and Boston.
Pulak, C. 1998. “The Uluburun shipwreck: an overview”, International Journal of Nautical
Archaeology 27:3, 188–224. DOI: 10.1111/j.1095-9270.1998.tb00803.x
Rankin, E. 1986. “Geometry enlivened: interpreting the refinements of the Greek Doric
temple”. Acta Classica: Proceedings of the Classical Association of South Africa,
número 29:1, 29-41. https://doi.org/10520/AJA00651141_416.
Rasmussen, C. 2015. “Clay model of a 1st century Roman War Ship” (available online https://
Rheeder, A. 2019. Architectural terracottas from Akragas: investigating monumental roofs
from the Archaic and Classical period (Unpublished doctoral dissertation), Leiden
Roberts, O. T. P. 1987. “Wind–power and the boats from the Cyclades”, International Journal
of Nautical Archaeology 16:4, 309–311. DOI: 10.1111/j.1095-9270.1987.tb00605.x
Rodríguez, F., ed. 1989–2019. Diccionario griego-español, Madrid.
Rose, C. B. 2013. The Archaeology of Greek and Roman Troy, Cambridge. DOI: 10.1017/
Roth, M. T., ed. 2010. The Assyrian Dictionary of the Oriental Institute of the University of
Chicago, Volume 20 U/W, Chicago.
Ruiz de Arbulo, J. 2009. “Los navegantes y lo sagrado. El barco de Troya. Nuevos
argumentos para una explicación náutica del caballo de madera”, in Arqueologia
Nàutica Mediterrània, X. Nieto and M. A. Cau, eds., Girona, 535–551.
Salt, A. M. 2009. “The Astronomical Orientation of Ancient Greek Temples”, PLoS ONE 4
(11): e7903. DOI: 10.1371/journal.pone.0007903
Sapirstein, P. 2016. “The Columns of the Heraion at Olympia: Dörpfeld and Early Doric
Architecture”, American Journal of Archaeology 120:4, 565–601. DOI: 10.3764/
Seidl, U. and W. Sallaberger. 2005–2006. ‘Der “Heilige Baum”’. Archiv für Orientforschung
Shiloh, Y. 1977. ‘The Proto-Aeolic Capital—The Israelite “Timorah” (Palmette) Capital’.
Palestine Exploration Quarterly 109(1): 39-52. DOI: 10.1179/peq.19188.8.131.52.
———. 1979. “The Proto-Aeolic Capital and Israelite Ashlar Masonry”, Qedem 11, 1–95.
Singer, I. 1988. “The Origin of the Sea Peoples and Their Settlement on the Coast of Canaan”,
in Society and Economy in the Eastern Mediterranean (c. 1500-1000 BC).
Proceedings of the International Symposium held at the University of Haifa from the
28th of April to the 2nd of May 1985, M. Heltzer and E. Lipiński, eds., Leuven, 239–
Spawforth, T. 2006. The Complete Greek Temples, London.
Taylor, R. 2005. “Roman Oscilla: An Assessment”, RES: Anthropology and Aesthetics 48,
Tóth, J. A. 2002. “Composite Stone Anchors in the Ancient Mediterranean”, Acta
Archaeologica Academiae Scientiarum Hungaricae 53:1–3, 85–118.
Throckmorton, P., ed. 1987. History from the Sea: Shipwrecks and Archaeology, London.
Thureau-Dangin, F. 1921. Rituels accadiens, Paris.
Ussishkin, D. 2013. “Sennacherib’s Campaign to Judah: The Events at Lachish and
Jerusalem”, in Isaiah and Imperial Context: The Book of Isaiah in the Times of
Empire, A. T. Abernethy, M. G. Brett, T. Bulkeley and T. Meadowcroft, eds., Eugene,
Votruba, G. 2017. «Did Vessels Beach in the Ancient Mediterranean? An assessment of the
textual and visual evidence». The Mariner’s Mirror 103 (1): 7-29. DOI:
Wachsmann, S. 1981. “The ships of the Sea Peoples”, International Journal of Nautical
Archaeology 10:3, 187–220. DOI: 10.1111/j.1095-9270.1981.tb00030.x
———. 1995. “Paddled and Oared Ships Before the Iron Age”, in The Age of the Galley:
Mediterranean Oared Vessels since Pre-Classical Times, R. Gardiner and J. Morrison,
eds., London, 10–35.
———. 2019. “On the Interpretation of Watercraft in Ancient Art”, Arts 8:4, 1–65.
Wallinga, H. T. 1995. “The Ancestry of the Trireme 1200-525 BC”, in The Age of the Galley.
Mediterranean Oared Vessels since Pre-Classical Times, R. Gardiner and J. Morrison,
eds., London, 36–48.
Way, A. S., trans. 1898. The Tragedies of Euripides vol. III, London and New York.
Wiggermann, F. A. M. 1992. Mesopotamian Protective Spirits. The Ritual Texts. Groningen:
Styx & PP.
Wilson Jones, M. 2002. “Tripods, Triglyphs, and the Origin of the Doric Frieze”, American
Journal of Archaeology 106, 353–390. DOI: 10.2307/4126279
———. 2014. Origins of Classical Architecture: Temples, Orders and Gifts to the Gods in
Ancient Greece, New Haven and London.
Wurz, R. 1914. Spirale und Volute von der vorgeschichtlichen Zeit bis zum Ausgang des
Altertums: mit besonderer Berücksichtigung des ionischen und des korinthischen
Kapitells, vol. I, Munich.
Yadin, Y. 1968. “‘And Dan, Why Did He Remain in Ships’ (Judges, V, 17)”, Australian
Journal of Biblical Archaeology 1:1, 9–23.
Yannopoulos, S. I., Lyberatos, G., Theodossiou, N.,Valipour M., Li W., Tamburrino, A. and
Angelakis, A. N. 2015. “Evolution of Water Lifting Devices (Pumps) over the
Centuries Worldwide”, Water 7, 5031-5060. DOI: 10.3390/w7095031.