THE NORTH END OF THE CITY

According to Kalokairinos: ‘An avenue leading from Herakleio to Knossos crosses the land of Knossos and is flanked by Roman ruins, built of mortar and small stones' (M. Kalokairinos 1906–7, 4; cf. Kopaka Reference Kopaka1989–90, 28). ‘c.1000 m’ north of the Kefala Hill (which is the hill of the Minoan palace) ‘is the tomb of Caiaphas … and two tombs of Hellenic date were found in 1881. The cemetery of the city of Knossos perhaps lies to the right of these tombs and is covered by 2–3 m of soil’. From this area up to the Kefala Hill

The monuments Kalokairinos describes have been destroyed or lie under the southern suburbs of modern Herakleio, on both sides of the Knossos Avenue. The ancient road leading to Knossos has, however, been sufficiently documented and there is evidence for monuments flanking it.Footnote ⁸ Kalokairinos locates the cemetery of Knossos by the tomb of Caiaphas, a Late Roman mausoleum located west of the Venizeleio hospital.Footnote ⁹ This monument had long been (erroneously) associated with the homonymous high priest of the New Testament, and was the prime tourist attraction in Knossos from the eighteenth century until its destruction in 1878.

Kalokairinos offers above two brief archaeological notes on the thick surface layers overlying the cemetery and on the discovery of two graves of Hellenic date. The Cretan scholar does not specify the type of the graves, but the vicinity of the tomb of Caiaphas later produced Early Iron Age chamber tombs (Hood and Smyth Reference Hood and Smyth1981, 38 nos 55–6). Roman graves are also known from this area and at least one of them was explored in the nineteenth century and yielded a golden statuette of a Nike (Halbherr Reference Halbherr1893, 112; Xanthoudides Reference Xanthoudides1901, 315; Chatzidakis Reference Chatzidakis1931, 19; Platon Reference Platon1955, 137).

Closer to (but still north of) the Kefala Hill, Kalokairinos identified two rows of stones with mortar on the left side of the visitor heading south. These are probably the ‘ruines Romains’ (sic) of the map in Fig. 2, and can be identified as the remains of the Roman Civil Basilica, the survival of which in primarily two rows is clearly identifiable in illustrations of the late nineteenth century.Footnote ¹⁰ Halbherr, who dug in the area in 1885, confirms the destruction of the Basilica by the Ottomans and calls this an ‘act of Vandalism’ (Halbherr Reference Halbherr1893, 111; cf. Morgan Reference Morgan2009). Building materials (especially marbles) from the Basilica and perhaps other Knossian monuments were used for the erection of the Ottoman barracks at Herakleio (Halbherr Reference Halbherr1893, 110–11; Xanthoudides Reference Xanthoudides1901, 305; Hood and Smyth Reference Hood and Smyth1981, 1). Indeed, the newspaper Ἀκρόπολις (23 July 1886, p. 3) reports that the locals protested at the removal of Corinthian capitals and pieces of geisa (cornices) from the Basilica for use in the barracks.Footnote ¹¹ The Ottoman building was erected in 1883 and took the place of the Venetian barracks of Saint George, which had been built in 1585 and had been destroyed by an earthquake in 1856. The Ottoman barracks can be seen at Eleftherias Square and currently house the regional administration of Crete and the law courts (Xanthoudides Reference Xanthoudides1927, 68, 127, 151–2).

The marbles from Knossos that were taken to the seat of the regional administration included a cuirassed statue of Emperor Hadrian, which belongs to a well-known sculptural type represented in different Cretan cities.Footnote ¹² Iossif Chatzidakis saw the statue standing in that building in 1881, but this was later donated to the Archaeological Museum of Herakleio and is currently on display in room XXVII (Fig. 3).Footnote ¹³ The Knossian provenance of the statue is established by the reference of Chatzidakis, and finds additional support in the museum catalogue and the published guide (Platon Reference Platon1955, 145 no. 5). Literature on Roman sculpture has missed this evidence and considers the piece to have been found at Gortyn.Footnote ¹⁴ The Knossian provenance is confirmed, however, by a little-known report by Richard Wyatt (Squire) Hutchinson, which establishes that the legs of the statue were discovered north of the Civil Basilica in 1935.Footnote ¹⁵ The overlooked testimonies by Chatzidakis and Hutchinson solve a confusion that has persisted for over a century, and establish that there are two similar statues of a cuirassed Hadrian from Knossos, and none from Gortyn.

Fig. 3. Statue of Hadrian from Knossos in the Archaeological Museum of Herakleio (inv. no. 5). Photograph by the author. © Archaeological Museum of Herakleio – Ministry of Culture and Sports – Archaeological Receipts Fund.

The references by Kalokairinos and Chatzidakis to the destruction and removal of Knossian antiquities are not isolated. Several foreign visitors from the sixteenth century onwards describe the systematic demolition of ancient structures in Knossos and the use of the building material in Herakleio (Kopaka Reference Kopaka, Cadogan, Hatzaki and Vasilakis2004, 498, 503–5, 508; also Xanthoudides Reference Xanthoudides1901, 305; Hood and Smyth Reference Hood and Smyth1981, 1; Paton and Schneider Reference Paton, Schneider and Chaniotis1999, 281–2; Paton Reference Paton, Cadogan, Hatzaki and Vasilakis2004, 453); one other such case is described below.

THE ACROPOLIS HILL AND THE CITY WALLS OF KNOSSOS

Kalokairinos comments:

Kalokairinos refers here to two monuments on the east and west side of the Acropolis of Knossos.Footnote ¹⁶ The theatre (or amphitheatre) on the lower east slope of the hill, which is also mentioned in his French manuscript and is demarcated on the accompanying map (Fig. 2) (Kopaka Reference Kopaka1989–90, 18), is discussed by visitors of the sixteenth, eighteenth and nineteenth centuries (Kopaka Reference Kopaka, Cadogan, Hatzaki and Vasilakis2004, 501–4; Moore Reference Moore2010, 22, 44). It must have been damaged in the 1880s when the modern road to Knossos was constructed (Hood and Smyth Reference Hood and Smyth1981, 42 no. 110; Paton Reference Paton, Evely, Hughes-Brock and Momigliano1994, 152; Paton and Schneider Reference Paton, Schneider and Chaniotis1999, 282), but part of the cavea is shown on maps of the late nineteenth and early twentieth centuries (Hogarth Reference Hogarth1899–1900: general map; Panagiotaki Reference Panagiotaki, Cadogan, Hatzaki and Vasilakis2004, 526–7, fig. 48.7). Sections of the monument were traced in rescue excavations in 1977 and 1992–3 (Sweetman Reference Sweetman2010, 344, 349, 360, 364).

Very interesting is the discussion of the circuit walls of Knossos. Kalokairinos introduces this with a passage from Strabo (10.4.7) (M. Kalokairinos 1906–7, 3–4; cf. Kopaka Reference Kopaka1989–90, 27): κύκλον ἔχουσα ἡ Κνωσσὸς τὸν ἀρχαῖον τριάκοντα σταδίων (‘Knossos has an ancient perimeter of 30 stadia [c.5.5 km]’). Strabo's reference has occasionally been taken to imply the existence of city walls, but this interpretation has been doubted (Hood and Smyth Reference Hood and Smyth1981, 20, 23, 36 no. 37; Coutsinas Reference Coutsinas2013, 358). Likewise, the reference of the eighteenth-century traveller Richard Pococke (Reference Pococke1745, 256) to ‘some little remains of the walls’ in the north part of the city has been associated with the Makriteichos, the retaining wall of a large Roman stoa (Hood and Smyth Reference Hood and Smyth1981, 23, 40–1 no. 86). Thus, current scholarship concludes that the city had no wall circuit (Hood and Boardman Reference Hood and Boardman1957, 226; Coutsinas Reference Coutsinas2013, 65–6, 357). Indeed, evidence for defensive structures at Knossos remains rare, and a Hellenistic round tower with two curtain walls forming an angle, which was discovered on the Kefala ridge, beyond the north edge of the ancient city, has been called ‘the first undisputed fortification of any period of antiquity’.Footnote ¹⁷ The monument was considered ‘more likely to belong to an isolated fort than to form a part of a city wall round the whole of Knossos' (Hood and Boardman Reference Hood and Boardman1957, 224, 226; cf. Coutsinas Reference Coutsinas2013, 65–6, 357–8, 421–2).

It is worth revisiting the widespread assumption about the paucity of archaeological evidence for defensive structures at Knossos in the light of the passage above, but also on the basis of an unpublished find from a rescue excavation of 1984 in the east end of the plot of the Venizeleio hospital.Footnote ¹⁸ The wall, which is c.7 m long and 1.60 m broad, and has a minimum height of 1.5 m, is aligned with (and perhaps related to) the fort on the summit of the Kefala ridge. It is made of a row of boulders overlying rows of roughly rectangular blocks, a style of masonry that recalls Archaic walls, including terraces, from elsewhere in Crete (Donald Haggis, pers. comm.). The pottery lying in the vicinity of the wall is all of Greek date.

Kalokairinos reports above on a much longer stretch, which ran for c.100 m along the west side/foot of the Acropolis of Knossos and may have formed part of a circuit wall, rather than of an isolated fort protecting the Acropolis.Footnote ¹⁹ The Cretan scholar certainly interpreted it as such, as evidenced by his map in Fig. 2. Kalokairinos discusses the size and construction of the monument, but the dimensions he gives are problematic. Stone blocks of 2.40 m in length are highly unlikely, and their reported breadth of 0.73 m cannot easily fit the very narrow breadth of the wall (1.20 m). Perhaps Kalokairinos confused the figures for the length of the individual blocks and the breadth of the wall, but this cannot be confirmed. In any case, the construction of this wall is dissimilar to that of the fort on the Kefala ridgeFootnote ²⁰ and the possible stretch east of the Venizeleio, and also seems different to Late Antique fortifications from elsewhere in Crete (Tsigonaki Reference Tsigonaki2007, 272–6).

The passage of Kalokairinos cited above explains the paucity of evidence for defensive structures in Knossos by describing extensive stone robbing. Similar episodes are documented in the nineteenth century and earlier (see above), but this case is interesting in that the Ottoman authorities intervened to prevent the complete destruction of the monument. It would be worth revisiting terraces and earth scarps on the west part of the Acropolis to see if any traces of the wall reported by Kalokairinos survive and can be investigated more fully with geophysics. The west part of the Acropolis Hill is poorly researched (Hood and Smyth Reference Hood and Smyth1981, 45–6, nos 149–50 and 152), but Todd Whitelaw informs me that the Knossos Urban Landscape Project (KULP) has traced a sharply defined fall-off of surface material at the south-west and along the west slope of the Acropolis in both the Prehistoric and Greek periods, which could be explained by the constraining effect of fortifications on the spread of occupation. This sharp edge of the sherd distribution can serve as a starting point for future investigations.

THE LOWER GYPSADHES HILL

Kalokairinos offers much new information on the archaeology of the Lower Gypsadhes Hill:

The hill, which is located by Kalokairinos at a distance of 300 m south of the Kefala, to the right of the road that leads south of Knossos, is the Lower Gypsadhes Hill.Footnote ²¹ The Cretan scholar reports on four marble statues and numerous terracotta statuettes found on this hill in 1873. Kalokairinos had first-hand knowledge of the statue of Theseus, which he sent to Athens, and the terracottas, which he acquired for his collection. It is, however, probable that he never saw the remaining statues, and that of the ‘son of king Phoenix’ was probably only shown to him in a photograph.

The Lower Gypsadhes Hill was explored on several occasions in the twentieth century, especially by David Hogarth in 1900, Humfry Payne in 1933, and Sinclair Hood, Peter Fraser and Nicolas Coldstream in 1957–60 (Coldstream Reference Coldstream1973b, 1a; Hood and Smyth Reference Hood and Smyth1981, 56, nos 286 and 297). More recently, the hill was surveyed by the KULP in 2005 and 2007, while its lower northern slope was explored by the Knossos Gypsadhes Geophysics Project in 2010–11, and excavated by the Knossos Gypsadhes Project in 2014–15.Footnote ²² The finds from the sanctuary of Demeter on the lower slopes of this hill best recall those mentioned by Kalokairinos in including hundreds of terracottas and several pieces of marble sculpture (Coldstream Reference Coldstream1973a); indeed, the site was known as ‘The Terracottas’ among the British excavators. The identification of the sanctuary was once attributed to Humfry Payne in 1927,Footnote ²³ but recent archival research has traced it back to the late nineteenth century. Indeed, Evans visited the site in 1896 and saw ‘a votive deposit of terracottas’ (Fig. 4) (Brown and Bennett Reference Brown and Bennett2001, 249–50). Closer to the time referred to by Kalokairinos (1873), two letters of 1879–80 written by Thomas Backhouse Sandwith, then British Consul at Khania, record his acquisition of hundreds of clay figurines from the same site.Footnote ²⁴

Fig. 4. Sketch map of Knossos by Sir Arthur Evans, from Notebook C, p. 94, 1896, Sir Arthur Evans Archive. © Ashmolean Museum, University of Oxford.

The looting of the sanctuary of Demeter on the Lower Gypsadhes Hill may have started as early as the sixteenth century, as Luigi Beschi has shown that a group of marble statues in Venice probably comes from this site (Beschi Reference Beschi1972–3, 494–9; cf. Hood Reference Hood1987, 87 n. 14). Kalokairinos attributed the looting of the 1870s to an unnamed man, but Sandwith specifies that this was the Muslim owner of the land, who employed several men to dig.Footnote ²⁵ Sandwith tried to sell the terracottas he acquired to the British Museum, but Charles Thomas Newton, then Keeper of the Department of Greek and Roman Antiquities (on whom see Sakellarakis Reference Sakellarakis1998, 207–8), suspected that these were forgeries and prevented the purchase. Nevertheless, there are specific terracottas in the Louvre, which Reynold Higgins identified as Cretan on the grounds of fabric and style and associated with the Knossian sanctuary of Demeter on the basis of iconography.Footnote ²⁶ This suggestion finds support in archival information. Several of the pieces identified by Higgins as Knossian were donated to the Louvre by Charles Simon Clermont-Ganneau,Footnote ²⁷ a French scholar who visited Crete in 1895 and met Kalokairinos. On that occasion, Kalokairinos donated a number of antiquities from his collection to the Louvre and actually invited Clermont-Ganneau to choose these pieces.Footnote ²⁸

The sanctuary of Demeter at Knossos can also be regarded as the most probable source for the considerable group of Cretan terracottas at the Museum of Leiden. These have previously been identified as Knossian on stylistic grounds, and are known to have reached the museum in 1882 through the collector Stephanos Savva Nikolaides of Smyrna and Richard Jacob van Lennep, the Dutch consul at this city.Footnote ²⁹ Kalokairinos also intended to export the figurines he acquired from Lower Gypsadhes and donate them to the National Archaeological Museum at Athens, but the finds were lost when the Ottoman mob looted and burnt the Kalokairinos mansion at Herakleio on 25 August 1898 (M. Kalokairinos 1906–7, 21; cf. Spanakis Reference Spanakis1960, 297–8; Kopaka Reference Kopaka1989–90, 10; 1995, 510; 1996, 151–2, 154–7; 2015, 146, 150 n. 5).

In the passage above, Kalokairinos proposes the identification of the figurines he possessed with images of Britomartis, which is doubtful. Many female, but also male and animal figurines were later excavated from the sanctuary of Demeter, but their association with particular divinities is not facilitated by any iconographic attributes. Indeed, the identification of female figurines of specific types with Demeter relies on the epigraphic documentation of her cult at the site (Higgins Reference Higgins1973, 61, 78; Coldstream and Higgins Reference Coldstream and Higgins1973, 184). Kalokairinos further reports that Ernst Fabricius studied the finds and divided them into three phases, but the German scholar did not include any relevant comment in his publication of several finds unearthed by the Cretan scholar.Footnote ³⁰ The tripartite classification is, however, reflected in the inventory of Kalokairinos's Knossian terracottas, which reports 45 clay heads of Britomartis of ‘archaic art’, 164 examples of ‘later art’, and 63 of ‘even later art’ (Kopaka Reference Kopaka1996, 155). Additionally, it records 25 body fragments of clay statuettes of the goddess, 17 pieces showing her on a throne, and 19 fragments of terracotta animals. Not all terracottas need to have come from the same findspot, but all are compatible with a provenance from the sanctuary of Demeter, which has yielded the only major deposit of figurines from Knossos. Additionally, the enthroned type was the commonest at this site and was also seen there by Evans (Higgins Reference Higgins1973, 77–80; Coldstream and Higgins Reference Coldstream and Higgins1973, 184; Brown and Bennett Reference Brown and Bennett2001, 249–51).

The sanctuary of Demeter was not equally rich in the kind of stone sculpture mentioned by Kalokairinos, but yielded one Hellenistic marble statue identified with Kore, and several fragments of marble figures, figurines and reliefs, in addition to three inscribed marble pieces, all of Hellenistic or Roman date.Footnote ³¹ Geoffrey Waywell explained: ‘That so few marble figures were found … merely testifies to the zealous activities of stone robbers and lime burners in this area in later times' (Waywell Reference Waywell1973, 97). Following the testimony of Kalokairinos (and Sandwith), one should add clandestine excavators to this list. Indeed, some of the numerous robbing pits and trenches, which Coldstream associated with ‘post-antique plundering’ of the site, may have been caused in 1873 (Coldstream Reference Coldstream1973b, 4–6, 13, 15; cf. Hood Reference Hood1987, 89). Two Roman inscribed monuments from Knossos mentioning Demeter and Kore, which were seen in Herakleio in the late nineteenth century, must have been looted from the sanctuary of Demeter.Footnote ³²

One of the four statues mentioned by Kalokairinos in the passage above may come from the sanctuary of Demeter. This is the marble statue of a figure 1 m in height, which, according to his description, held a duck and was inscribed ‘son of king Phoenix’ (Φοίνικος βασιλέως υἱός). Classical mythology knows of two figures by the name king Phoenix, a Greek and a Phoenician (RE 39, 404–14; OCD ⁴, 1140). The Greek figure is Phoenix, petty king of the Dolopes and elderly advisor of Achilles in the Iliad. Because of wrongdoings in his youth, Phoenix remains childless and does not have any son, as the inscription from Knossos requires. He, however, calls Achilles his son (Homer, Iliad, 9.492–4). The second Phoenix is the son of Agenor of Phoenicia and brother of Cadmus, Cilix and Europa. He settled in north Africa (Hyginus, Fabulae, 178; Antoninus Liberalis, Metamorphoses, 40) and had three sons, Cilix, Phineus, and Doryclus, and a stepson, Atymnius, fathered by Zeus (Scholia on Apollonius Rhodius's Argonautica 2.178; cf. Wendel Reference Wendel1958, 140). Some treated Phoenix as the father of Cadmus (Scholia on Apollonius Rhodius's Argonautica 3.1186; cf. Wendel Reference Wendel1958, 252), and others as the father of Europa (Homer, Iliad 14.321; Euripides, fragments 472 and 752 g in Collard and Cropp Reference Collard and Cropp2008). The Phoenician dynasty is closely associated with Crete through Europa, but also through Phoenix, who is mentioned among other heroes in the Hellenistic inscription of the oath of the ephebes from Dreros (ICr I ix 1, line 30; cf. Capdeville Reference Capdeville2014, 154–5) and is considered as the father of Itanos, oikistes of the homonymous east Cretan city (Stephanus of Byzantium, s.v. Itanos; cf. ICr III, p. 76; Capdeville Reference Capdeville2014, 151–2). Nevertheless, neither the Greek nor the Phoenician Phoenix can be readily associated with the statue in question, and especially with the inscription and the iconographic attribute of the duck.

A ‘son of Phoenix’ (Φοίνεικος υἱόν) is mentioned also in the transcription of a Roman inscription from Knossos, which was shown to Federico Halbherr by a dealer of antiquities in Herakleio in the 1880s or early 1890s. The transcription includes 19 more letters, which are hard to understand: Φοίνεικος υἱὸν εἰσορᾶ|ς ΣΕΝΕΚΓΩΝΑΜΕΣ.Footnote ³³ The text was reconstructed by Arthur Maurice Woodward as a pair of tetrameters (with a metrical error): Φοίνεικος υἱὸν εἰσορᾷς Σενεκ(ί)ωνα Μέσ[σιον] (‘You are looking upon Senekion Messios, son of Phoenix’) (Woodward Reference Woodward1936, 95; cf. Baldwin Bowsky Reference Baldwin Bowsky1999, 480, 485 no. 20; 2006, 420 no. 6). The Phoenix in question is not a king, and has been tentatively identified as a ‘familiaris’ of the procurator P. Messius Campanus (Baldwin Bowsky Reference Baldwin Bowsky1999, 480). This reconstruction is incorrect, as shown below.

It is most likely that the description provided by Kalokairinos and the transcription obtained by Halbherr pertain to the same monument, even if the match is not perfect: the name Phoenix is spelled slightly differently; the transcription has ‘son’ in the accusative (υἱόν), but Kalokairinos gives it in the nominative (υἱός); and the reference to a king is missing from the transcription. These discrepancies can be attributed to problems in recording. The transcription, which was shown to Halbherr and was not made by this fine epigrapher, has previously been listed among ‘copie errate’ and has been characterised ‘non omnino rectum’ (Ricci Reference Ricci1893, 305 no. 14; ICr I viii 32). Additionally, Kalokairinos probably only saw the monument in a photograph and published his notes on it more than three decades after that viewing, in 1906. Problems of accuracy have previously been raised with respect to other inscriptions mentioned by Kalokairinos (Kopaka Reference Kopaka1996, 156 n. 9).

These problems make the question of the survival of the inscribed statue all the more pressing. A search for the name Φοίνειξ in the Searchable Greek Inscriptions of the Packard Humanities Institute yields only one further attestation (<http://epigraphy.packhum.org/search?patt=φοίνει>). Although the text of this ‘second’ inscription is very similar to the one from Knossos, the provenance given for the piece is Egypt. The two monuments have hitherto been treated as distinct, but they should be identified as one. Indeed, in the text cited above, Kalokairinos reports that he tried to purchase the Knossian inscribed statue for the National Archaeological Museum in Athens, but this was smuggled to Alexandria.Footnote ³⁴ This monument was published as soon as it reached Egypt, as early as 1873, in an anonymous report in the Zeitschrift für ägyptische Sprache und Altertumskunde.Footnote ³⁵ The report attempts the laundering of the statue right in its first line: ‘Long ago, a statue was found in Lower Egypt (currently held in the house of a Swiss)’ (Anonymous 1873, 127). The statue had just arrived in the country and needed to be naturalised; its Cretan provenance was written off and was hence forgotten for almost a century and a half. The Swiss collector mentioned is an obscure figure, as confirmed to me by Vassilis Chrysikopoulos, who has investigated the history of collections in Egypt.

The published report describes an inscribed Greek statue 70 cm tall. ‘Made in good style, the statue shows an unclothed boy, who holds a goose or a duck in his left hand, and an elongated vase in his right hand. The vase stands on an altar-like base, which carries the following inscription copied with precision: Φοίνει|κος υἱ|ὸν εἰσ|ορᾶς Σε|νεκιῶ|νά με’ (Φοίνει|κος υἱ|ὸν εἰσ|ορᾶς Σε|νεκίω|νά με in Preisigke Reference Preisigke1915, 105 no. 1165) (‘You are looking upon me, Senekion, son of Phoenix’). The text adds that the inscription was read by Heinrich Karl Brugsch, a well-known German Egyptologist, and was assigned to the Late Roman period. Clearly, Brugsch's reading of the inscription is much more convincing than that offered by Kalokairinos, by the recording in the transcription from Herakleio, and also by the reconstruction proposed by Woodward.

The identification of the inscribed statue confirms that Kalokairinos was incorrect in reading ‘king’ in the inscription and suggests the figure depicted was not any mythological character. How could Kalokairinos make such an addition? There is reason to believe that this was more than a casual error. It is indicative that Odysseus Kalokairinos called the legendary king Minos ‘the son of Phoinix’ (a title otherwise unattested for the king, who is called only the grandson of Phoenix in some sources; see above) and also considered him as ‘a Jew from Phoenicia’ (O.M. Kalokairinos Reference Kalokairinos1939, 7, 94; cf. Karademas Reference Karademas, Andreadaki-Vlazaki and Papadopoulou2011, 823). Minos Kalokairinos was convinced that the Minoans were Semitic in origin, as is clearly manifested in a discussion he had with Baron Nathan Mayer Rothschild. In 1886, Kalokairinos hosted the British banker and politician, showed him his collection of antiquities and took him to Knossos, which makes the Baron the first ‘VIP’ to tour the site (O.M. Kalokairinos Reference Kalokairinos1939, 47, 49–50). The Cretan scholar described to the Baron (who was of the Jewish faith) his view on the Semitic origins of the Minoans and explained that this remained unknown, probably because of anti-Semitism, a point that stimulated a broader discussion on economics and politics. It was probably his quest for evidence confirming the migration of the Phoenician ruling dynasty to Knossos that directed Kalokairinos to the erroneous reading of the inscription.

Notwithstanding the reading by Kalokairinos, the two names recorded in the inscription are personal names. Both are otherwise unattested in Crete, but Φοίν(ε)ιξ is quite common elsewhere in the Greek world (e.g. LGPN 1, 475), whereas Σενεκίων is very rarely attested (only in LGPN 1, 403 and LGPN 5A, 401). Inscriptions recording a name and patronymic and addressing the passer-by are typically found on tombstones.Footnote ³⁶ Indeed, the term εἰσορᾷς is commonly found on Greek funerary monuments and occurs on two examples from west Crete.Footnote ³⁷ On this basis, one may assume that the inscribed statue comes from one of the monumental Roman tombs found in the area of the sanctuary of Demeter and Hogarth's houses,Footnote ³⁸ or from further south, beyond the summit of the Gypsadhes, where Hellenistic and Roman built tombs with inscribed blocks were (reportedly) found in the early and mid-twentieth century (Hood and Smyth Reference Hood and Smyth1981, 59 nos 326, 327, 334).

Although the inscription clearly favours a funerary provenance, the iconography of the statue does not exclude a different possibility. Indeed, the piece in question conforms to a sculptural type that was common in the Hellenistic and Roman periods and was represented in both cemeteries and (predominantly) sanctuaries.Footnote ³⁹ In her study of the Hellenistic examples, Olympia Bobou demonstrated that statues of boys holding offerings, especially ducks or other birds, are particularly fitting for sanctuaries, including those of Demeter and Kore.Footnote ⁴⁰ This raises the possibility that the statue from Knossos comes from the sanctuary on the Lower Gypsadhes Hill, which yielded a few Hellenistic terracottas with comparable iconography (Higgins Reference Higgins1973, 86 no. 229; 75 nos 121 and 130). Nonetheless, there is considerable chronological discrepancy between the terracottas and the marble statue, whose provenance from a burial context is more likely on epigraphic grounds.

More reliable is the identification of the find context of the statue of Theseus mentioned by Kalokairinos in the passage above. Indeed, his informant suggested that it was found ‘on top of this hill’ (i.e. the Lower Gypsadhes Hill), a location that was also visited by Evans, Hogarth and Halbherr at the end of the nineteenth century. Evans saw an unfluted column and Roman walls and foundations on the spot (Fig. 4; Brown and Bennett Reference Brown and Bennett2001, 249–51). A few years later, Hogarth identified (but did not discuss) the remains of a Roman heroon in precisely this location.Footnote ⁴¹ There is some information on this monument in Hogarth's unpublished diary.Footnote ⁴² On 19 March 1900, the British scholar noted: ‘Finished the “Heroon” + measured it up + photoed it.’ This is the only explicit reference to the heroon in his diary, but the entries on previous days contain possible references to the same monument. On 14 March, Hogarth identified ‘remains of some building fallen to ruin near top centre [of the Lower Gypsadhes Hill] + site has been so much cultivated that Mycenaean and Roman pottery is all one on top of other’. On 16 March, he noted ‘High up [on the hill] a wall showing of good masonry but Roman? No finds but three ghost coins (one, Nero: one province), a bronze lance head + a pierced clay button. A large basin with three bands rude moulding round rim unearthed near top of hill.’ Lastly, on 17 March 1900 he recorded: ‘Halbherr came out in morning and thought oblong building on summit Hellenic.’

According to these descriptions, the building on top of the Lower Gypsadhes Hill can be reconstructed as well-built, oblong and adorned with columns. The finds were of mixed date, but Roman material predominated, even if Halbherr was inclined to raise its chronology and, more recently, Peter Callaghan has argued for a Hellenistic date (in Hood and Smyth Reference Hood and Smyth1981, 22–3, 58, no. 312; cf. Alcock Reference Alcock2002, 122). The drawing(s) of the building that Hogarth mentions could not be traced and the collection of photographs from his fieldwork that is kept in the archives of the British School at Athens does not illustrate any monument that can be identified with the heroon. The uncertainties over this monument have discouraged any discussion of the identity of the hero venerated, but Martha Bowsky has hypothesised that a cult of the Dioskouroi was housed there (among other Knossian locations) (Baldwin Bowsky Reference Baldwin Bowsky2006, 399). The find context of the inscribed statue of Theseus, as documented by Kalokairinos, provides important evidence in this respect.

Structures, cults and rituals honouring Theseus are rarely attested in Classical antiquity. The relevant evidence is concentrated in the city of Athens and is even missing from rural Attica (Walker Reference Walker1995, 20–4). The single exception known to me concerns the sacrifice that the Athenians offered to Theseus at Cape Rhium in Achaia, where they dedicated a ship captured in a sea battle of 429 bc (Pausanias 10.11.6; cf. Icard-Gianolio Reference Icard-Gianolio, Lambrinoudakis and Balty2004, 270–1). Crete has previously yielded no such evidence (cf. Sanders Reference Sanders1982, 36–40; Sporn Reference Sporn2002), but the mythology of Theseus connects him specifically to Knossos. Indeed, Susan Alcock has made a case for the rise of the cults of Minos and the characters associated with him (specifically, but tenuously, including Theseus) in Roman Crete. She identifies this as a new pattern of commemoration, which involved a diminution in the power of the stories about the past that were only known locally, in favour of celebrated stories that could be recognised by audiences across the empire (Alcock Reference Alcock2002, 123–30, 179–80). Additionally, the Cretan labyrinth, the venue of Theseus's labour, is often mentioned in Roman literature, and Philostratus (Vita Apollonii 4.34) localises this monument in Knossos and presents it as an attraction pointed out to visitors like Apollonius of Tyana and his companions. A heroon of Theseus, perhaps visually or spatially associated with the assumed location of the labyrinth, would be particularly fitting to Roman Knossos.

The Roman date of this monument is confirmed by the rediscovery of the statue of Theseus mentioned above by Kalokairinos. According to the Cretan scholar, the statue of Theseus was sent to Philippos Ioannou and the National Archaeological Museum in Athens. The relevant correspondence of the two men has not been found, but the statue has been traced to the Epigraphic Museum at Athens (no. 2815) and appears in the records of the Archaeological Society at Athens and the National Archaeological Museum at Athens, where it is specifically identified as a gift by Minos Kalokairinos.Footnote ⁴³ What is left of the monument is the base preserving the feet of the statue and the inscription Θησέως (Fig. 5). Kalokairinos recorded Θησεύς, but the difference is slight and his documentation of inscriptions was not flawless (see above and below). In fact, the material (marble), the size (half life-size) and the fragmentary state of preservation confirm the identification. The statue was found in Knossos in 1873, was registered in the inventory of the Archaeological Society at Athens in 1875 and was first published in 1881 by Ludwig von Sybel, who saw it at the Varvakeio Lykeio in 1879–80.Footnote ⁴⁴ The identification of the monument is valuable in establishing the hitherto unknown provenance of the piece.Footnote ⁴⁵ The date in the first or second century ad deduced for the statue on epigraphic criteria accords with the chronology suggested by Hogarth and Evans for the heroon (as opposed to the earlier date proposed by Halbherr and Callaghan).

Fig. 5. Statue of Theseus in the Epigraphic Museum in Athens (no. 2815). © Epigraphic Museum, Athens.

Significantly, the earliest description of the piece from the Epigraphic Museum includes a reference to a fragment of a right arm holding a club, a weapon that is commonly held by Theseus in Roman art.Footnote ⁴⁶ As Frank Brommer observes, there is no trace of the club on the base and this must have been held against the shoulder, as in the case of the Theseus shown on a wall painting from Pompeii.Footnote ⁴⁷ The same position is attested on a Roman statue from west Crete, which was found in the lararium of an urban villa at Kissamos. The young, beardless hero with the thin club from Kissamos was identified as Herakles by Stavroula Markoulaki (Reference Markoulaki, Loukos, Xifaras and Pateraki2009, 357 fig. 12), but the resemblance between Theseus and Herakles in Roman art is well known.Footnote ⁴⁸ Further assumptions on the iconography of the statue from Knossos and its dependence on well-known sculptural types, including the Hermes Ludovisi and Ares Borghese, are insecure.Footnote ⁴⁹

The findspot of the two uninscribed statues that the landowner of the Lower Gypsadhes Hill mentioned to Kalokairinos cannot be identified with any precision. They may be among the statues found later at the sanctuary of Demeter, or could come from the Roman tombs that are located in the area, as noted above. Another possible provenance is indicated by the discovery of a small Hellenistic marble statue of a Nike south of the bridge over the Vlychia stream (Waywell Reference Waywell1973, 97–8; Hood and Smyth Reference Hood and Smyth1981, 56 no. 287). However, the piece was found in the ravine and may have eroded from the adjacent slopes.

The fate of most of the finds from Gypsadhes recorded by Kalokairinos was unfortunate. The terracottas must have perished in the fire of 1898, along with the rest of his collection. However, as Kopaka has observed, it is unlikely that all those antiquities were destroyed without leaving a trace (Kopaka Reference Kopaka, Macdonald, Hatzaki and Andreou2015, 150 n. 5). It is worth considering whether some of the probably Knossian terracottas in European museums were once held by Kalokairinos and were looted from his mansion in 1898. The fate of the two uninscribed statues is uncertain and they remain to be relocated. It is unlikely that Kalokairinos purchased them, although his collection included many marble statues from unspecified Knossian locations.Footnote ⁵⁰ Lastly, the statues of Theseus and of the ‘son of Phoenix’ were exported outside Crete, in accordance with a practice that can be traced back to the sixteenth century (Beschi 1972–3, 494–9; Hood Reference Hood1987, 87; Paton and Schneider Reference Paton, Schneider and Chaniotis1999, 281; Kopaka Reference Kopaka, Cadogan, Hatzaki and Vasilakis2004, 503). Kalokairinos sent the statue of Theseus to Athens in at least two fragments, but only the base survives and is presently located at the Epigraphic Museum. The Cretan scholar tried to purchase the statue of the ‘son of Phoenix’ for the National Archaeological Museum at Athens, but this was smuggled to Alexandria, where it was studied and published as early as 1873 as coming from Lower Egypt and located ‘in the collection of a Swiss’. To my knowledge, the monument has not resurfaced since and its current whereabouts remain unclear.

THE TOMBS AT SPILIA

According to Kalokairinos:

The Roman rock-cut tombs of Spilia (Figs 1 and 2) were known to visitors since the early nineteenth century (if not earlier) and these monuments and adjacent quarries were occasionally identified with catacombs or with the Knossian labyrinth.Footnote ⁵¹ Halbherr provides further information on the visit of 1892, though without mentioning Kalokairinos.Footnote ⁵² The Italian scholar describes many plundered tombs and notes he excavated one (Halbherr Reference Halbherr1893, 112; cf. Paton Reference Paton, Evely, Hughes-Brock and Momigliano1994, 149). He also comments on the inscription mentioned by Kalokairinos:

There can be little doubt that Kalokairinos and Halbherr are describing the same inscription, which has remained unknown because the letters were poorly preserved and the monument was stolen shortly afterwards. From the testimonies of the two scholars we can deduce that this was a funerary inscription threatening tomb robbers with a fine; it was written in Greek, but dated from the Roman period.Footnote ⁵³

Kalokairinos suggests that the inscription mentioned a certain Fiscos, but no figure by that name is known from Cretan history and this name is not recorded in the LGPN.Footnote ⁵⁴ It therefore comes as no surprise that Halbherr avoided this reference. Instead of a proper name, one should read here the masculine noun φίσκος (from Latin fiscus), which designates the imperial treasury, and should identify a standardised provision threatening potential tomb violators with liability to the treasury.Footnote ⁵⁵ Many of these monuments date to the second and third centuries AD, but there are also later examples.Footnote ⁵⁶ A search for the term φίσκος in the Searchable Greek Inscriptions reveals that inscriptions with the provision in question are found across much of the ancient world, but are common only in Asia Minor (<http://epigraphy.packhum.org/search?patt=φισκ>), as Halbherr (Reference Halbherr1893, 112) suggested. The Italian scholar considered the Knossian inscription as unique for Crete and this has remained so after over a century of research, but Martha Bowsky kindly informs me that there is an unpublished comparable monument from Lissos.

The misreading of the word fiscus by Kalokairinos and his misunderstanding of other inscriptions discussed above raises questions over the accuracy of the remaining few words he documented, particularly since Halbherr resisted any reconstruction of this text. The title of the ‘archon of the Romans’ is peculiar and its reliability is undermined by the erroneous addition by Kalokairinos of the comparable term ‘king’ in the inscription of the ‘son of Phoenix’. The Greek term archon can render different Roman magistrates (Mason Reference Mason1974, 111–13), but, to my knowledge, the designation ‘archon of the Romans’ does not appear in any of the inscriptions mentioning the provision in question (<http://epigraphy.packhum.org/search?patt=φισκ>), and magistrates are generally not recorded in this kind of monument (Lempidaki Reference Lempidaki2015, 188–9). However, many of these monuments, especially those from Asia Minor, mention the terms ‘Caesar's treasury’ (Καίσαρος φίσκον / Καίσαρος φίσκῳ) or ‘the lord's treasury’ (κυριακὸν φίσκον / κυριακῷ φίσκῳ), which are synonymous. It is probably the latter of the two terms that the Cretan scholar translated as ‘archon of the Romans’. Taking together the various hints offered by Kalokairinos and Halbherr and drawing from the rich evidence on this epigraphic type, one can reconstruct the Knossian inscription as having: a clause against those entering more bodies in the tomb; the standardised provision ἀποτ(ε)ίσει/δώσει εἰς τὸν κυριακὸν φίσκον or ἀποτ(ε)ίσει/δώσει τῷ κυριακῷ φίσκῷ (‘to deposit in the lord's treasury’); and an indication of the amount due. Tomb inscriptions recording threats of fines payable to the imperial treasury have been connected indirectly to the imperial cult (S.R.F. Price Reference Price1984, 119), for which there is epigraphic evidence at Knossos (Sanders Reference Sanders1982, 38; Paton Reference Paton, Cadogan, Hatzaki and Vasilakis2004, 453; Baldwin Bowsky Reference Baldwin Bowsky2006, 400; Lagogianni-Georgakarakos Reference Lagogianni-Georgakarakos, Cadogan, Hatzaki and Vasilakis2004).

KNOSSIAN ANTIQUITIES IN PRIVATE HANDS

Kalokairinos notes:

In this passage, Kalokairinos singles out two (or three?) ancient objects from his excavations in the area of the Minoan palace. The first object is called a coin or a relief and the second object, which is called a statuette, has proved to be a relief. This confusion was probably caused by the three decades that elapsed between Kalokairinos's excavation and the composition of this passage. The silver coin that shows Europa on a tree belongs to a Gortynian numismatic type of the Classical and Hellenistic periods (Svoronos Reference Svoronos1890, 167–9, pl. XII; ICr IV, p. 37–8). Such a precious find is unlikely to have reached the Kefala accidentally and may have been dedicated to the temple of Rhea, which was built on top of the South Propylaeum of the Minoan palace and yielded more silver coins of the Early Classical period (Evans Reference Evans1928, 5–7, 346, 349; Coldstream Reference Coldstream2000, 285–8, 296; Prent Reference Prent, Cadogan, Hatzaki and Vasilakis2004, 416–18). Kalokairinos is assumed to have dug up votive offerings from the Rhea sanctuary, perhaps including a bird vase of the Middle Geometric period that was recorded by Fabricius (Reference Fabricius1886, 142, pl. 3: top left; cf. Hood and Smyth Reference Hood and Smyth1981, 32 n. 73; Coldstream Reference Coldstream2000, 288–9 no. K1) but disappeared in the destruction of the Kalokairinos mansion.

The second object described by Kalokairinos as ‘a statuette of a seated female’ has been identified by Kopaka with a Roman relief of the first or second century ad (Fig. 6) on the basis of the acquisition no. 15 in the Archaeological Museum of Herakleio.Footnote ⁵⁷ The relief is usually taken to show Demeter seated on the agelastos petra or mirthless stone, but the figure is also holding a cornucopia. There are a few more stone reliefs showing Demeter seated, but these are much earlier (fourth century bc), come from her sanctuary at Eleusis and show considerable differences in iconography.Footnote ⁵⁸ None of these works and no other illustration of a seated Demeter/Ceres has the goddess holding a cornucopia.Footnote ⁵⁹ This attribute is more fitting to Tyche, who is shown seated on a rock on a relief of the second or first century bc.Footnote ⁶⁰ Kopaka is sceptical about the provenance of the Knossian relief from the excavation of Kalokairinos, given the paucity of Roman finds from the Minoan palace.Footnote ⁶¹ Nonetheless, Chatzidakis (Reference Chatzidakis1881, 15) reports the discovery of ‘parts of marble columns and pieces of statues’ in the excavation of Kalokairinos, and more recent fieldwork has traced Hellenistic and Roman finds, including a sanctuary deposit, on the north-west periphery of the palace.Footnote ⁶²

Fig. 6. Relief of a goddess from Knossos in the Archaeological Museum of Herakleio (inv. no. 15). Photograph by the author. © Archaeological Museum of Herakleio – Ministry of Culture and Sports – Archaeological Receipts Fund.

Beside the two finds, the passage above mentions three contemporaries of Kalokairinos. One of them is his field director, the schoolmaster Christos Papaoulakis, who was immortalised in the award-winning novel of Rhea Galanaki, Ο Αιώνας των Λαβυρίνθων (The Century of the Labyrinths, 2002). The second person is the owner of the land, Zekiris Beji Ibrahim Efedakis, who is also known from Ottoman property contracts (Panagiotaki Reference Panagiotaki, Cadogan, Hatzaki and Vasilakis2004, 516–17). The last person is Ioannis Fotiades, governor of Crete in 1878–85, who intervened and stopped Kalokairinos's fieldwork in Knossos in 1879.Footnote ⁶³ Although this intervention was in accordance with a decision by the Cretan Assembly, it makes particularly interesting the contrast that the Cretan scholar draws between his own archaeological interests and those of Fotiades. Both men exported antiquities from the Ottoman province of Crete to Greece. However, Kalokairinos donated these pieces to the National Archaeological Museum of Athens through the aid of Philippos Ioannou, Professor of Philology at the University at Athens, whereas Ioannis Fotiades sold Cretan antiquities to another Professor of the University of Athens, Athanasios Rhousopoulos, who was a well-known dealer.Footnote ⁶⁴ Rhousopoulos had a reputation as a collector of coins (Galanakis Reference Galanakis2011, 172, 192), so the silver piece that Fotiades acquired from Kalokairinos's dig may have reached his collection, which included numerous examples of this coin type (Hirsch Reference Hirsch1905, 176–7, nos 2970–83).

THE KNOSSIAN PERIPHERY: HERAKLEIO, ARCHANES AND JUKTAS

According to Kalokairinos:

Although very brief, the testimonies of Kalokairinos on Herakleio and Archanes are particularly interesting. The finds from Herakleio were reported in the local press in 1886 as the first archaeological evidence for the existence of an ancient city on the spot.Footnote ⁶⁵ The coastal location of these tombs is important, since all other Greek and Roman burials known from Herakleio are located in the city centre (Karetsou Reference Karetsou2008, 57–63; cf. Kotsonas Reference Kotsonas2002, 64). Stephanos XanthoudidesFootnote ⁶⁶ confirmed the coastal location of these tombs at Bedenaki, but dated their finds to the Roman period. Amabile Ittar (1830?–1904), the man who donated the finds from these tombs to the local museum, was a doctor and vice-consul of Italy, France and Austria. He was personally acquainted with Kalokairinos and occasionally appears in the correspondence of archaeologists active in Crete at the time (Detorakis Reference Detorakis1994; also Spanakis Reference Spanakis1960, 298; Christofi Reference Christofi1996, 365; Brown and Bennett Reference Brown and Bennett2001, 249; Panagiotaki Reference Panagiotaki, Cadogan, Hatzaki and Vasilakis2004, 513, 519; Vakirtzian Reference Vakirtzian2006, 406–8). The mansion of the Ittar family survives at modern Herakleio in Epimenidou Street (number 18).

The Juktas wall was well known by the nineteenth century and dates from the second millennium bc, but its precise chronology is uncertain.Footnote ⁶⁷ To my knowledge, the second, long stretch of wall located north of Archanes is otherwise undocumented.Footnote ⁶⁸ However, Kalokairinos also commented on it (and on the Juktas wall) in the lecture he delivered at the palace of Knossos in 1903:

The two texts disagree on the length of the wall and the location. The Journal suggests a location to the north of Archanes, whereas the lecture mentions toponyms to the west and south of the site. The syntax in the second text is, however, problematic and it is most likely that Kalokairinos cited the two toponyms with reference to the assumed temple, rather than the wall. Chosto Nero, a small cave on the west slope and by the south peak of Juktas, has produced Middle Minoan figurines and Hellenistic and Roman material (Faure Reference Faure1964, 175–6; Sakellarakis and Sapouna-Sakellaraki Reference Sakellarakis and Sapouna-Sakellaraki1997, 43, 68–9, 323, 512, 514). Agios Mamas is a chapel located on the steep slope of a low hill on the south end of Epano Archanes and is unknown in archaeological literature. To my knowledge, neither of the two locations preserves evidence of a fortification wall, and it is hard to envisage such a monument at either of the two.