ArticlePDF Available

In Search of the Pillars of Heracles



The Pillars of Heracles were columns ancient Greeks regarded as marking the boundaries of the furthest west. In Greek mythology, Heracles laid down the columns when sailing to the island Erytheia at the edge of the world. It is argued the original columns were located during the time of Homer (late 8th century BCE) and Hesiod (c. 700 BCE) at Epirus (northwestern Greece) but were relocated to the south Iberian Peninsula in the 630s BCE, when Greeks expanded their geographical horizon and began trading with Tartesso-Iberians.
Spring 2019 Vol. I.III
Table of Contents
Letter from the Editors
Jake Freeman: The Dangers of the East: To what extent is Arrian’s anabasis a
critique of growing orientalism under the Antonine Emperors?
Astrid Khoo: Reconceptualising the Classics through Shakespeare’s Plays
Justin Muchnick: µένδέ: Modernization and Museumification in the New Acropolis
Isabella Riglia: Byzantine culture outside the boundaries of its empire: An inquiry into the
exportation and adoption of culture in the Middle Byzantine period
Holly Smith: Reconstructing the Making Processes from Material Acquisition to Finished
Object of the Bronze Bull and Bull-Leaper Group and a Terracotta Bull Figurine
Oliver D. Smith: In Search of the Pillars of Heracles
Dear Readers,
In this our third issue of Kerberos, we are excited to present you with six new and intriguing papers by
undergraduate classicists. In our commitment to the wide scope of “Classics” we have carefully chosen essays
that explore those somewhat more niche areas of the discipline for our summer issue.
This issue takes us away from the more traditional topoi of Classics, exploring Bronze age Making, Byzantine
culture, Roman orientalism, and more contemporary discussions on Classical receptions and Museum debates.
We hope to help guide our readers through an exploration of the multiple facets of the classical discipline and
perhaps aide in broadening the scope of your conception of “Classics”.
On a more personal note, this will be our final issue as Editors. We are overjoyed with the reception of our little
journal and have been deeply moved by the support of our peers and our mentors as we have navigated Kerberos
inception and establishment within the King’s College London Classics Department! We want to thank everyone
who has pushed us to keep working, despite a few speed-bumps along the way. Thank you to our wonderful
team of editors, we could not and frankly would not have done this without you. Thank you to our readers, for
continuing to interact with us online and keeping us on schedule with your comments and suggestions.
Thank you once again to Eris for being our muse.
A final thank you, and a warm welcome to Becky Brown and Emma Bentley! They will be
taking over from us as editors in the upcoming 2019-2020 academic year and we could not be
more excited to see what they bring to our “child” Kerberos!
It’s been a labour of love.
Bryony Callaghan & Giulia Vidoni
Editors in Chief
In Search of the Pillars of Heracles
Oliver D. Smith
Online version:
In Search of the Pillars of Heracles
Oliver D. Smith, Open University
The Pillars of Heracles were columns ancient Greeks regarded as marking the boundaries of the furthest west. In
Greek mythology, Heracles laid down the columns when sailing to the island Erytheia at the edge of the world.
It is argued the original columns were located during the time of Homer (late 8th century BCE) and Hesiod (c.
700 BCE) at Epirus (northwestern Greece) but were relocated to the south Iberian Peninsula in the 630s BCE,
when Greeks expanded their geographical horizon and began trading with Tartesso-Iberians.
The Pillars of Heracles were two columns or statues
located at the western edge of the world
to mark “furthermost limits of seafaring” reached by Heracles (Pind. Nem. 3. 23). In Greek
mythology, one of Heracles’ labours involved him travelling to the extreme west and sailing
Okeanos (the encircling ocean) to the island Erytheia, home of Geryon’s cattle (Hes, Th. 288-
293; Ps-Apollod. 2. 5. 10). Heracles either put down the columns on Erytheia (Schol. Pind. Ol.
3. 44) or the westernmost promontory of land before sailing Okeanos (Dio. Sic. 4. 18. 4). While
Heracles’ wanderings were likely imaginary – geography of his twelve labours involved real
similar to the itinerary of Odysseus’ wanderings and voyage of the Argonauts; the
columns were considered to delimit the utmost west by ancient Greek sailors (Strab. 3. 5. 5-6).
At least one source describes them as statues (Eustat. ad Dion. Perieg. 64) but they were more commonly thought
to be columns or pillars (stêlai). Strabo (3. 5. 6) notes they originally were man-made objects, but overtime they
crumbled and were replaced by an appellation with natural rock formations (such as mountains) that were nearby.
Dueck (2012: 25-27) discussing geography of the labours notes that: “Myths often gained a touch of authenticity
by including real geographical data”. Six (or seven) labours took place in the Peloponnese: Nemean lion (Argolis),
Lernaean Hydra (Argolis), Ceryneian golden-horned deer (Arcadia), Erymanthian boar (Arcadia), Augean stables
(Elis), Stymphalian birds (Arcadia) and Heracles entered the underworld Hades at Tarnarum (Laconia) to capture
Cerberus; one in Crete: Cretan bull (Knossos); one in Thrace: Diomedes’ Mares (Abdera); the three remaining
labours took place in faraway places at edges of the world (Geryon’s cattle, Amazons and apples of Hesperides).
Columns in Iberian Peninsula
The earliest classical source to mention the columns, Hecataeus’ Periodos ges (frg. 6 Müller)
written in the late 6th century BCE places them at Mastia, traditionally identified as Cartagena,
Spain. Hecataeus seems to have been alone in locating the columns at Mastia; the geographer
Pseudo-Scymnus (142-146) placed them near Mainake, that archaeologists now identify with
the city Málaga (about 320 km west of Cartagena). In contrast, the 5th century BCE astronomer
Euctemon thought the Pillars of Heracles were on the opposite sides of the Strait of Gibraltar;
one column on Kalpe, meaning Gibraltar’s promontory; another at Mount Abilyx
(Euct. ap.
Av. Ora. 333-340). The renowned historian Herodotus in The Histories (c. 440 BCE) positions
a column on the Libyan coast (4. 43. 3-4, 181. 1, 185. 1; 8. 132. 1) and claims Tartessos
near the Pillars of Heracles (4. 152. 2) which vaguely puts the columns opposite the same strait.
There was disagreement about the precise location of the Libyan column at the Strait; the 1st
century BCE geographer Strabo (3. 5. 5) notes while it was sometimes identified with Mount
Abilyx (Erat. frg. 106 Roller) a different theory placed it on an islet (perhaps the island Perejil).
Ignoring the ancient dispute about the Libyan column, Strabo (3. 5. 5) informs us most Greeks
generally accepted Heracles’ columns were in the vicinity of the Strait of Gibraltar, citing for
example Dicearchus, Eratothenes and Polybius; to this list can be added: Pseudo-Skylax (1),
(Ti. 24e), Pseudo-Apollodorus (2. 5. 10) and Dionysius Penegetes (64-73). Posidonius
(frg. 246 Kidd) however rejected this more popular view and argued for a separate tradition:
the Pillars of Heracles sat inside a Punic temple dedicated to Melqart at Gádeira (Cádiz), Spain.
Monte Hacho (Ceuta) or Jebela Musa (Morocco) both opposite of Gibraltar.
Tartessos was a region usually identified by ancient Greeks with the Guadalquivir valley in the southern Iberian
Peninsula, but Eratosthenes (frg. 153 Roller) argued “Tartessis” covered a broader territory reaching to the strait.
Plato’s description of a “mouth” near the Atlantic Ocean (and columns) can only mean the Strait of Gibraltar.
Figure 1. Four locations of the Pillars of Heracles in the south Iberian Peninsula.
Based on archaeological discoveries, it is known Greeks established commercial contact with
Tartesso-Iberian tribes on the southern and southeast Iberian Peninsula during the last third of
7th century BCE
(Domínguez, 2006: 436). This date is corroborated by Herodotus (4. 152. 2)
who records a tradition of a merchant from Samos, named Kolaios, who discovered Tartessos
in the 630s BCE. A simple solution to the four locations of the columns (i.e. Mastia, Mainake,
Strait of Gibraltar and Cádiz, see Figure 1.) is they moved when Greeks increased geographical
knowledge of the Iberian Peninsula from the southeast coast; westward, beyond the Strait: “the
Pillars later had to be moved to accord with the fact the Straits had become penetrable” (Romm,
1992: 17). In Greek myth, Erytheia that sat just beyond the columns, was placed near Tartessos
(Stes. ap. Strab. 3. 2. 11), Gádeira (Hdt. 4. 8. 2), the Strait (Erat. frg. 153 Roller) or an isle off
Libya; Ptolemy’s Geography (c. 150 CE) identifies Erytheia with the Moroccan Mogador Isle.
The reluctance of some Greeks to re-locate the columns from the Strait to Gádeira might have
stemmed from hostile Carthaginian activity, or lack of accurate topographical information. In
regard to the former, Strabo (17. 1. 19) and Pseudo-Aristotle’s On Marvellous Things Heard
(84) mention Carthaginians tried to prevent Greeks sailing beyond the columns, while its likely
A small number of Greek pottery sherds have been discovered in the south of the Iberian Peninsula that predate
the 7th century BCE by one or more centuries, however, as Domínguez (2006: 432-433) cautions, they show signs
of having been transported by Phoenicians and can’t be used as evidence for Greek contact with Tartesso-Iberians.
Carthaginians spread fictitious stories to Greeks about waters beyond the Strait as being poorly
navigable to dissuade their rival trading competitors from going there.
Herodotus (4. 196. 1)
had heard tales about Carthaginian exploration beyond the columns but was always hesitant to
relocate the Pillars of Heracles from the Strait, to Gádeira (Cádiz); he was apparently unable
to find a first-hand eyewitness who had seen the Atlantic Ocean (3. 115. 2) beyond the columns.
Columns in Central Mediterranean?
It is tempting to argue that before ancient Greeks located the Pillars of Heracles in the Iberian
Peninsula, that they did so outside of Spain, when their geographical horizon was more limited.
The poet Hesiod who flourished c. 700 BCE had heard of the island Erytheia, writing Heracles
sailed Okeanos to reach it (Th. 289-294). Noticeably, Hesiod doesn’t localise Erytheia in the
south Iberian Peninsula, unlike Stesichorus’ Geryoneis, written a century later (Curtis, 2011:
15). While Hesiod doesn’t provide any geographical clues (only that Erytheia lay in Okeanos,
somewhere in the far west
), the fact he predated discovery of the Iberian Peninsula by more
than half a century means he couldn’t have thought Erytheia sat there, nor Heracles’ columns.
A non-Iberian location for the Pillars of Heracles has elicited the interest of modern Atlantis
enthusiasts, who seek to find the island of Atlantis described by Plato beyond the columns (Ti.
24e). Evidence clearly points to Atlantis being an imaginary island (Smith, 2016) and very few
classical scholars take seriously the notion Atlantis existed, despite this view remains popular
among laypeople and amateur archaeologists. Although Plato rather unambiguously locates the
island of Atlantis in the Atlantic Ocean Atlantis enthusiasts who recognise the impossibility
of an Atlantic site for Atlantis (given the fact there is no geological or archaeological evidence
for an ancient island civilisation) argue he must have been mistaken. The island and columns,
Water beyond the columns was chaotic (Schol. ad Pind. Ol. 3. 44) and too shallow with mud (Arist. Met. 2. 1).
Erytheia (from erythrós = red) translates as “red island” a reference to the western sunset.
they maintain were somewhere else, for example central Mediterranean.
An obvious problem
with this theory is there are no classical sources that place the Pillars of Heracles in this region.
In 2002, Italian journalist Sergio Frau published a book locating Heracles’ columns at the Strait
of Sicily citing a fragment attributed to Timaeus (c. 300 BCE) by Karl Müller: “The island of
Sardinia, near the Pillars of Heracles” (frg. 28). Despite Frau (2002: 285) arguing Timaeus had
written these words, if he read Müller’s Fragmenta Historicorum Graecorum he would realise
the source is On Lycophron (796) written in the 12th century by Byzantine poet John Tzetzes.
The mistake was made by Müller to list this as a fragment of Timaeus (when Tzetzes doesn’t
quote or reference Timaeus); it was subsequently removed in the Fragmente der Griechischen
Historiker. Anello (2014) discusses the context of Tzetzes placing the columns near Sardinia
and notes what he had in mind was “projection of Sardinia towards limit between civilization
and barbarism”. In other words, Tzetzes’ comment shouldn’t be taken literal, but as a metaphor.
Another Atlantis enthusiast who has erroneously said there are classical sources that position
the columns at the Strait of Sicily (stretching to the Lesser Syrtis) is Anton Mifsud, author of
the book The Pillars of Heracles: Myth or Reality? (2017). None of the sources Mifsud cites
such as Apollonius Rhodes (4. 1230 f.) actually place the columns in this region; the classicist
Vidal-Naquet (2007: 126) has noted recent attempts to re-locate the Pillars of Heracles to the
Sicilian Strait are poorly justified since no “single ancient text” supports such an identification.
Columns in Epirus
As it happens there do exist some classical sources that put Erytheia outside of the south Iberian
Peninsula, but not in the central Mediterranean. A fragment of Hecataeus’ Genealogies (late
6th century BCE) preserved by the 2nd century CE historian Arrian (An. 2. 16. 5) says the home
This argument is traceable to an article “Atlantis – A New Theory” (Weir, 1959) that claims the columns were
originally located at the Strait of Messina, however, the Strait of Sicily remains by far the more popular hypothesis.
of Geryon, where Heracles’ stole his cattle, was in the area between Ambracia and Amphilochi,
on the southern boundary of Epirus (Greece). A separate Epirotic tradition recorded by Pseudo-
Skylax (26) locates Erytheia nearby, but to the north, at Kestria. According to Fowler (2013:
299) Hecataeus possibly extended Geryon’s kingdom northward into Thesprotia. Furthermore,
ancient Greeks thought cattle in Epirus and surrounding territories descended from Geryon’s.
Although some classicists have argued Hecataeus’ localisation of Geryon or Erytheia in Epirus
was a rationalisation of the story, others consider the Epirotic tradition to be the earliest when
the geographical horizon of the Greeks was more limited when the utmost west was Epirus and
the Ionian Sea (northwestern Greece). This is supported by reconstructing Homer’s (late 8th
century BCE) and Hesiod’s (c. 700 BCE) geographical knowledge, that establishes the Ionian
Sea as Okeanos, in the remote west. Odgen (2014: 2012) points out “there was a time when the
Thesprotian coast, the westernmost edge of the mainland occupied by the Greeks, was regarded
as a sort of ‘ultimate west’.” Arguably, Homer (by a few decades) predated the earliest Greek
colonies in Italy and Sicily, which is why he was unfamiliar
with the central Mediterranean.
Homer vaguely knew of Sicels across the Ionian Sea (Od. 24. 211) but he seems to have had no topographical
information about Sicily: “To the west he has heard vaguely of the tribes of southern Italy and Sicily, the Sikeloi
and Sikanioi, but they are mere names to him” (De Camp, 1970: 217). Eratosthenes (frg. 6 Roller) claimed Homer
had no geographical knowledge of the central Mediterranean region. Hesiod had heard of places such as Mount
Etna because he was contemporary to the earliest Greek colonies in Italy and Sicily, but his knowledge was hazy,
rather than accurate; the latter is for example demonstrated by his error Etruria was a group of islands (Th. 1015).
Anello, P. (2014). “Tradizioni etnografiche e storiografiche sulla Sardegna (Diod. IV 29-30;
82; V 15)”. Hormos: Ricerche di Storia Antica. 6: 1-20.
Curtis, P. (2011). Stesichoros’s Geryoneis (ed., trans.) Mnemosyne supplements. Monographs
on Greek and Latin language and literature, 333. Leiden: Brill.
De Camp, L. S. (1970). Lost Continents. New York: Dover Publications.
Domínguez, A. J. (2006). Greeks in the Iberian Peninsula” in Tsetskhladze, C. R. (ed.) Greek
Colonisation: An Account of Greek Colonies and Other Settlements. Leiden; Brill. 429-506.
Dueck, D. (2012). Geography in Classical Antiquity. Cambridge: Cambridge University Press.
Fowler, R. L. (2013). Early Greek Mythography: Volume 2. Oxford: Oxford University Press.
Kidd, I. G. (1999). Posidonius: Volume 3. Cambridge: Cambridge University Press.
Mifsud, A. (2017). The Pillars of Heracles: Myth or Reality?. Prehistoric Society of Malta.
Ogden, D. (2014). “How ‘Western’ Were the Ancient Oracles of the Dead?” in Breglia, L.,
Moleti, A. (eds.) Hespería: Tradizioni, Rotte, Paesaggi. Paestum: Pandemos. 211-226.
Romm, J. S. (1994). The Edges of the Earth in Ancient Thought: Geography, Exploration, and
Fiction. Princeton, NJ: Princeton University Press.
Smith, O. D. (2016). The Atlantis Story: An Authentic Oral Tradition?. Shima: The
International Journal of Research into Island Cultures. 10(2): 10-17.
Roller, D. W. (2010). Eratosthenes’ Geography. Princeton, NJ: Princeton University Press.
Vidal-Naquet, P. (2007). The Atlantis Story. Exeter: University of Exeter Press.
Weir, A. R. (1959). “Atlantis – A New Theory”. Science Fantasy. #35 (June).
ResearchGate has not been able to resolve any citations for this publication.
This paper considers the Mycenaean metallurgical industry at the end of the Bronze Age through analysis of metal hoards and the tools found within them. An overview of second-millennium hoards from Crete and the Greek mainland is presented to contextualize the various objects from these assemblages. Patterns of implement inclusion reveal a repeated tool grouping in seven Mycenaean hoards, most associated with elite contexts. These Mycenaean caches, incorporating a range of complete and broken items, are traditionally considered recyclable scrap, but they need not be random accumulations. The repetitive tool grouping suggests a structural principle in hoard formations, perhaps dictated by the state. The Mycenaean assemblages from the late 13th or early 12th century B.C.E. highlighted here may represent either palatial stock within a citadel or an archaeological remnant of a metal ta-ra-si-ja allotment given to a smith. The identification of several hoards as metal disbursements from a palace complements the picture of Pylian administration of metal, attested only by the ta-ra-si-ja arrangements outlined in the Jn and Ja Linear B tablets.