Content uploaded by Manolis Manoledakis
Author content
All content in this area was uploaded by Manolis Manoledakis on Jan 23, 2025
Content may be subject to copyright.
Great Waterworks in
Roman Greece
Aqueducts and Monumental
Fountain Structures
Function in Context
edited by
Georgia A. Aristodemou and
Theodosios P. Tassios
Archaeopress Roman Archaeology 35
Archaeopress Publishing Ltd
Gordon House
276 Banbury Road
Oxford OX2 7ED
www.archaeopress.com
ISBN 978 1 78491 764 7
ISBN 978 1 78491 765 4 (e-Pdf)
© Archaeopress and the authors 2018
Cover: The monumental arcade bridge of Moria, Lesvos,
courtesy of Dr Yannis Kourtzellis
Creative idea of Tasos Lekkas (Graphics and Web Designer, International Hellenic University)
All rights reserved. No part of this book may be reproduced, in any form or
by any means, electronic, mechanical, photocopying or otherwise,
without the prior written permission of the copyright owners.
Printed in England by Oxuniprint, Oxford
This book is available direct from Archaeopress or from our website www.archaeopress.com
i
Contents
Preface ���������������������������������������������������������������������������������������������������������������������������������������������iii
Georgia A. Aristodemou and Theodosios P. Tassios
Introduction I� Roman Aqueducts in Greece �������������������������������������������������������������������������������������1
Theodosios P. Tassios
Introduction II� Roman Monumental Fountains (Nymphaea) in Greece �����������������������������������������10
Georgia A. Aristodemou
PART I: AQUEDUCTS
Vaulted-roof aqueduct channels in Roman Macedonia ������������������������������������������������������������������15
Asimina Kaiafa-Saropoulou
The aqueduct of Actian Nicopolis ����������������������������������������������������������������������������������������������������26
Konstantinos L. Zachos and Leonidas Leontaris
The water supply of Roman Thessaloniki ����������������������������������������������������������������������������������������50
Manolis Manoledakis
The Hadrianic aqueduct of Athens and the underlying tradition of hydraulic engineering ���������70
Eustathios D. Chiotis
The Hadrianic aqueduct in Corinth �������������������������������������������������������������������������������������������������98
Yannis Lolos
The Roman aqueduct of Mytilene �������������������������������������������������������������������������������������������������� 109
Yannis Kourtzellis, Maria Pappa and George Kakes
Roman aqueduct of Samos ������������������������������������������������������������������������������������������������������������� 131
Τelauges Ν. Dimitriou
A Roman aqueduct through the Cretan highlands – securing the water supply for elevated Lyttos � 147
Amanda Kelly
PART II: NYMPHAEA
Shifting tides: approaches to the public water-displays of Roman Greece ����������������������������������173
Dylan Kelby Rogers
Fountain figures from the Greek provinces: monumentality in fountain structures of Roman
Greece as revealed through their sculptural display programs and their patrons ���������������������� 193
Georgia Aristodemou
The monumental fountain in the Athenian Agora: reconstruction and interpretation���������������218
Shawna Leigh
New water from old spouts: the case of the Arsinoe fountain of Messene ����������������������������������� 235
Mario Trabucco della Torretta
Reflecting the past: the nymphaeum near the so-called Praetorium at Gortyn ���������������������������246
Brenda Longfellow
ii
50
The water supply of Roman Thessaloniki
Manolis Manoledakis
Abstract
The city of Thessaloniki was so much increased in terms of both size and population in the Roman period that a new significant
water supply construction was urgently needed. This need was covered by a huge ambitious aqueduct that carried water from
Mt Kissos (modern Hortiatis) to the acropolis of Thessaloniki, and from there to the city quarters. It consisted of a series of
underground tunnels driven deep inside a mountainside to tap its aquiferous stratum, and a monumental bridge, of a total
length of ca. 20km. The aqueduct was constructed in the Roman period, but experienced several repairs and reconstructions
in the Early Christian, the Byzantine and the Ottoman periods, as can be assumed by careful observations and research. In this
paper we deal with the history of the Hortiatis aqueduct by examining the several construction phases of the monument as well
as the references to it in the written sources.
Keywords: Thessaloniki, Hortiatis, aqueduct, qanat, springs
Introduction: the city’s water needs
The location chosen by Cassander to build in 316/5 BC the new city, which he named after his wife, was
not accidental. Besides the fact that it was at the cove of a bay within the Thermaic Gulf, and therefore
offered a perfect and protected harbour, it also covered an area surrounded by an important mountain,
Hortiatis, then called Kissos, and several streams that flew very close to its walls.1
During the first phase of the history of Thessaloniki, the needs for water of its inhabitants, who, as we
learn from Strabo (7 frg. 21, 24) came from twenty six neighbouring small towns of which we know six
(Apollonia, Chalastra, Thermi, Gariskos, Aineia, Kissos), were covered by the area’s groundwater,2 which
was drawn from various wells scattered throughout the city. Evliya Çelebi mentions in the chapter of
the eighth book of Seyahatnâme (1667) that Thessaloniki had 3060 such wells in his time. This figure is
certainly exaggerated as far as the ancient times are concerned, but in any case, the Hellenistic city
seems to have had water autonomy; little more can be said on this issue, as very few parts of this early
city phase have been archaeologically researched.
In the Roman period, however, the situation changed. Already in the 1st century AD, Strabo (7 frg. 21)
calls Thessaloniki ‘metropolis of Macedonia’, while a century later Lucian (Lucius or the Ass 46) would call
it ‘the largest city of Macedonia’. It seems that the city experienced a remarkable increase of population,
as well as a consequent increase of its territory. It is no coincidence that it was chosen by Apostle Paul
as one of the cities he visited in his second missionary journey. Besides, the city flourished significantly
during the Imperial period, especially from the 1st century AD, when large public buildings appear, such
as administrative buildings, baths, the theatre-stadium, probably also the first phase of the Agora.3 It
is obvious that the needs for water increased significantly and the city’s groundwater was no longer
sufficient to cover them.
As a result, the inhabitants of Thessaloniki turned to other solutions, and the one that was offered as
particularly suitable was certainly the water supply of the city from the water sources of Mt Kissos
For these streams see Dimitriadis 1983: 218-221.
Cf. Lioutas 2000: 39, who estimates that the population of the Hellenistic city was no larger than 5000 people; Tamiolakis 2001: 13.
Vitti 1996: 58; Allamani-Souri 2003: 81; Adam-Veleni 2003: 143, 147, 155-158; 2004: 226.
51
(modern Hortiatis). Raised 1201m to the east of Thessaloniki, this mountain has always been famous for
its verdant environment and abundant waters. Many authors of all historical periods, from Xenophon
(Hunt. 11.1) until(De exp. Thess. 5), Evliya Çelebi (Seyahatnâme, book 8) and later travellers,
described in more or less detail and with a glorifying mood the fertile plains of the mountain, its dense
vegetation, its rivers and springs; During the winter large pieces of ice were formed on its numerous
lakes, which ‘resemble the eye of life’, and were sold during the summer by the people of the mount in
the city of Thessaloniki, even until the last century.4
All this indicates that Mt Hortiatis was an ideal solution for the city’s water supply problem, a solution
that was indeed soon used. The aqueduct of Hortiatis was built in the Roman period and for many
centuries supplied the city of Thessaloniki as well as other neighbouring settlements with water. Both
the route it followed and part of the construction itself are today known and visible near the entrance
of the village Hortiatis. However, until recently, some questions remained unanswered – and some still
need to be investigated – such as: the dating of its first construction phase, the dating of its repairs and
subsequent construction phases, which sector of Thessaloniki it supplied with water, and whether it was
earlier, contemporary or posterior in comparison with the other two aqueducts, those of Retziki and
Lembet, that are also reported as having supplied Thessaloniki.5 However, for the last two, much less is
known. We will examine all these separately.
The fact is that the excavations in ancient Thessaloniki have revealed several pipelines of water supply,
which are divided into three types: the built water pipes, the clay pipes, and the lead pipes.6 Most of them
were found in districts outside the ancient city, such as Toumba, Thermi, and Polichni.7 These water
pipes present two difficulties: the first relates to their routes, which, because of the small part that is in
all cases preserved, cannot be safely restored and thus identified with the course of either one of those
three aqueducts, or be regarded as a part a fourth one.8 The second difficulty concerns the precise dating
of the earliest of these pipes, which are generally dated to the Late Roman and/or the Early Christian
period.9 The presence of both water pipes and cisterns grows significantly, especially in the Upper Town
of Thessaloniki, during the late Early Christian and the Byzantine periods.10
There is also an honorary inscription of the 2nd century AD from Thessaloniki,11 which has been completed
by P. Nigdelis in a manner suggesting that it was referring to the repair of a city water supply system
from the area of Altia; the scholar connects Altia with the recorded in the sources -but still unidentified
– location named Altos.12 As Nigdelis points out, Altos cannot be connected to one of the three known
aqueducts, but the inscription is an important testimony to the water supply of Thessaloniki from distant
areas, already from the Imperial period.
Finally, it is suspected that the name ‘Caracalla’, which until the early 20th century was born by a region
of Pylea, between Hortiatis and Thessaloniki, could refer to an aqueduct built in that region by Emperor
Caracalla (AD 198-217),13 but this cannot be proven, as there are doubts even for the very fact of Caracalla
passing through the region of Thessaloniki.14
For the environment of Mt Kissos and its dramatic change since Antiquity, see Manoledakis 2007: 85-91.
Tamiolakis 2001; Lioutas 2000; Nigdelis 2006.
Cf. Lioutas 2000: 40, with relevant images.
Eleftheriadou 1992: 424; Makropoulou 1992: 424-425; Bulletin de correspondence hellénique (=BCH) 122, 1998: 871; Doulgeri 1998; Soueref 2000: 35;
Lioutas 2000: 40-46; Nigdelis 2006: 68.
Indicative is the case of the duct in Toumba, for which Eleftheriadou (1992: 24) says that it is Late Roman and irrelevant to the Hortiatis
aqueduct, while Doulgeri (1998: 221) considers it Early Christian and having Mt Hortiatis as its starting point.
See above, notes 7 and 8.
Nalpantis 1991.
IG X 2 1, 226; Bulletin de correspondence hellénique (=BCH) 46, 1922: 527.
Nigdelis 2006: 65-70. For Altos, which is mentioned by Stephen of Byzantium (s.v. Altos) to be a village near Thessaloniki, according to
Theagenes, see Cramer 1828: 238; Dimitsas 1874: 252; Papazoglou 1988: 203; Papakonstantinou-Diamantourou 1990: 104; 1997: 352. The fact that
Altos is placed on some maps in the region of Philyron (e.g. on the map in Papakonstantinou-Diamantourou 1997: 354), is due to its arbitrary, as
Papakonstantinou-Diamantourou (1997: 352), correctly points out, placement there by Dimitsas (1874: 252), although there is no evidence for
this placement.
Dimitriadis 1983: 423.
Allamani-Souri 2003: 71, 86-87.
Great Waterworks in Roman Greece
52
But let us see what we know about these three aqueducts of Thessaloniki.
The aqueducts of Retziki and Lembet
The aqueduct of Retziki starts near the homonymous village, so consequently it also carries water from
Mt Hortiatis. The name of the region is a corruption of a Turkish noun meaning paradise, a name given
precisely due to the abundant water and lush vegetation of the area. The water was transported initially
through water abstraction galleries and afterwards in closed pipes, covered with stone plaques and
coated internally with kourasani, and entered Thessaloniki passing through the Litaia Gate.15 From there
it ended up in a reservoir at the Monastery of the Twelve Apostles (today Church), which is dated from
the Byzantine period.16 G. Tamiolakis, the only one so far that has written in detail about this aqueduct,
considers that it may be older than that of Hortiatis, which he dates to the early 4th century. However, not
only is there no evidence to confirm such a view, but on the contrary the following data can contradict it:
1. All parts of the Retziki aqueduct that survive today, such as that in the area of Asvestochori, belong to
the Byzantine period; in the best case, some may possibly date in the Early Christian period.17 2. The dating
of the Hortiatis aqueduct in the 4th century AD has now been reversed, since new data, as we shall see in
the following paragraphs, date the first construction phase of the monument in the 1st century AD. 3. The
Retziki aqueduct was known as ‘New Water’ (Yeni Su)18 and was mentioned as such in contradistinction
to the Hortiatis aqueduct, suggesting that it was later than this.
The aqueduct of Lembet, like that of Retziki, supplied with water the western part of Thessaloniki.
Starting from the homonymous area (modern Stavroupolis), it ended up in the reservoir of the Twelve
Apostles church.19 This aqueduct is considered a construction of the Byzantine period,20 although parts
of water pipes in Stavroupolis and Polichni dating to the Roman times have recently been discovered,
which could be linked to this aqueduct.21 In the Ottoman period it was restored by the Grand Vizier of
Rumeli, . It should be noted however that in the very few – and all of a later period
– texts that mention this aqueduct, there is a confusion of information or even an identification with
information referring to the aqueduct of Retziki, as sometimes it is the Lembet aqueduct that is called
Yeni Su, and associated with the reservoir of the Twelve Apostles church.22 This could mean either a
connection of the two aqueducts, which were anyway situated very close to each other, or a confusion of
the scholars probably because of this very neighbourhood.
The aqueduct of Hortiatis
It is certainly remarkable that none of these two aqueducts, of Retziki and Lembet, was mentioned by any
medieval source or even by a modern traveller of Thessaloniki until the 19th century.23 In stark contrast,
as we shall see later on, the aqueduct of Hortiatis is mentioned in a variety of texts from as early as the
12th century. Moreover, the Hortiatis aqueduct provides several archaeological data that allow both a
more credible than in the other two aqueducts dating of its first construction phase and the pin-pointing
and dating of its subsequent repair phases. This is due to the two large parts of the aqueduct preserved
today, namely the qanat of Haghia Paraskevi and the part of the aqueduct bridge at the entrance of the
Hortiatis village, to both of which separate chapters will be dedicated here.
For the Retziki aqueduct, see Tamiolakis 2001: 29-35, who is cited by all the later authors writing about the aqueduct.
Hatzi Ioannou 1880: 66; Papageorgiou 1901: 38-39; Orlandos 1940: 377-383.
Tamiolakis 2001: 31-33, figs. 7-9.
Hatzi Ioannou 1880: 66.
Papageorgiou 1901: 38-39.
For the Lembet aqueduct, see Dimitriadis 1983: 426-432; Lazaridis 2001; Tamiolakis 2001: 36-39. According to data from the Ottoman period,
the amount of water from Lembet that reached Thessaloniki in this era was greater than that of the Hortiatis waters; however the latter
watered a bigger part of the city. See Dimitriadis 1983: 426, 433.
Lioutas 2000:45.
Tafrali 1913: 118-119; Letsas 1961:146; Spieser 1984: 13.
Cousinéry (1831: 106-107) refers to ‘Urendgik’, but only to connect it geographically with Mt Hortiatis (which he erroneously calls Disoron),
to the mountain range of which it really belongs, and without mentioning any separate aqueduct of this mountain.
53
The route to Thessaloniki and the city’s water supply
The part of the aqueduct that we can see today near the entrance of the Hortiatis village (Figures 1 and 6)
is a very small piece of a huge construction of approximately 20km, which started from the homonymous
mountain and ended up in the city’s reservoirs. The area where the waters of the mountain were collected
and started their route – initially underground – is the region of Haghia Paraskevi, near the homonymous
chapel to the east of the village of Hortiatis, on the northwest slopes of the mountain at an altitude of 575-
585m. A qanat was constructed there,24 the largest among others in Hortiatis,25 to collect the mountain
water (Figures 2-5). It is a water abstraction construction with a history of many centuries, a few samples of
which survive today in Greece. Some hundred meters further to the west of the qanat, water was exiting the
soil surface and led towards the city of Thessaloniki on an aqueduct bridge,which is preserved today near
the entrance of the Hortiatis village and is the monument that is today known as ‘the aqueduct of Hortiatis’
(Figure 1). This is a unique monument of this kind to be preserved in Central Macedonia. Afterwards, the
water pipe goes underground again, with short over ground intervals, depending on the natural terrain of
the area through which it passed, so that the desired inclination was maintained in any case.
See in detail below.
For the other qanats of Hortiatis, see Manoledakis 2007: 98, 116 and figs. 35 and 47.
Figure 1. Panoramic view of the aqueduct bridge of Hortiatis.
In the upper right side there is the entrance of the homonymous village.
Great Waterworks in Roman Greece
54
The route of the Hortiatis aqueduct, as this is indicated by several, much smaller, parts of it that survive
scattered between Hortiatis and Thessaloniki, shows that it entered the city from its so-called eastern part.26
After the aqueduct bridge of the village, the water route continued towards northwest, to the junction
of the roads to Asvestochori and Kavala, then towards Panorama; afterwards it turned northwest again,
to the area of the modern ‘Philippeion’ Hotel, from where it sloped down to the Citadel of Thessaloniki,
where a reservoir was situated.27 From there, the Hortiatis water entered the city, where it was distributed
to fountains, baths and reservoirs through a number of branches. The most important of these reservoirs
seems to have been at the Monastery of Vlatadon, which provided a large part of the city with water.28
Another branch of the aqueduct, which sent water to the eastern part of the city, has been identified in the
area of the Church of the Presentation of the Saviour (Ypapanti).29 Other branches are also mentioned; some
of them are probably earlier than the Vlatadon one; for example, the one that passed along the eastern
wall, ended up at the Gate of Kalamaria and brought water in a reservoir behind the Church of Haghios
Nikolaos Orfanos,30 or the one passing outside the eastern walls, in the area of the Aristotle University of
Thessaloniki.31
According to a view, the water of the Hortiatis aqueduct passed also from the Gate of Anna Palaiologina,
to serve other reservoirs and fountains of the city, such as those at the Church of Haghios Georgios, and
the fountain to the north of the Church of Haghia Sofia.32
The qanat
Description
The qanat is a very simple and effective way to tap the underground aquiferous stratum and transport
water with the help of the natural inclination to the level of the settlement. The term qanat that has
prevailed in the international literature is Arabian (in Persian it is called kariz). The word comes from the
Assyrian word qanu, which means reed – from which comes the synonymous ancient Greek word κάννα
(and through it the Latin canna) – and consequently whatever has the shape of a reed, i.e. pipe, duct.33
The qanat is a system of underground tunnels and pipelines, with the help of which the aquiferous stratum,
whose water is transported to the surface with the help of the slope gradient, is tapped.34 Consequently, the
settlement should be located close to a natural slope. In the upper part of the slope, the mother (or searching)
shaft – as it is called by Polybius (10. 28. 2) and Xenophon (Hell. 3. 1. 7) – is opened, in order to find the spring.
From this highest point of the qanat the tunnels and (in greater depths) the closed pipes begin to carry water
to the settlement, which has to be located on a lower level than that of the spring. Their gradient – necessarily
smaller than that of the slope so that they gradually reach the surface – usually does not exceed 5‰, or 0.5m
per km, the height of the tunnels ranges from 1.00 to 2.5m, their width from 0.75 to 1.5m, and their length from
a few meters up to 50 km. The water often flows in a channel, while all the remaining space serves the need to
be accessible by people for inspection, repairs, cleaning and generally maintenance.
In the international literature the qanat is considered a Persian (according to the prevailing view) or an
Armenian invention, which, based on the existent archaeological evidence, is dated no later than the 8th
In fact this is the southeastern part.
Cf. Letsas 1961: 145; Dimitriadis 1983: 424; Spieser 1984: 13; Tamiolakis 2001: 24. Cf. also Cousinéry 1831: 106-107.
Dimitriadis 1983: 423, 433; Stogioglou 1971: 52-53, who argues that this specific water supply belongs to an era earlier than the founding of the
monastery. For other branches of the aqueduct in Thessaloniki, cf. also Tafrali 1913: 117-118; Letsas 1961: 145.
Dimitriadis 1983, 61: 433.
Hatzi Ioannou 1880: 66. For the 26 branches of the water pipes of the Hortiatis aqueduct in Thessaloniki in the Ottoman period see
Dimitriadis 1983: 432-442. Cf. Tamiolakis 2001: 41-51.
Hatzi Ioannou 1880: 66; Tamiolakis 2001: 46-49.
Tafrali 1913: 117; Letsas 1961: 145; Tamiolakis 2001:41. This view is not accepted by Dimitriadis (1983: 433).
One can also meet the terms foggara (in North Africa) and madjira, after which Madrid was named (the city of qanats). Cf. Hodge 1992: 392
note 8; Vavliakis (1989: 1) and Wulf (1968: 94) etymologize, not convincingly, the term qanat from the ancient Semitic word gano (dig), as they
do also with the word channel.
The most important study of the qanat system is that of Goblot (1979), with exhaustive bibliography. For additional bibliography, see Hodge 1992:
391 note 4. Also recently, Tölle-Kastenbein 1990: 39-42; Hodge 1992: 20-24; Grewe 1998: 33-40 and passim; Hodge 2000: 35-38; Schacht 2003: 411-423.
In the Greek bibliography the work of Vavliakis (1989; 1997; 2000) almost monopolizes. In the 11th century the Arab mathematician Mohamed Al
Karagi wrote a book in Persia, recording details about the qanat construction techniques. See, Grewe 1998: 33-40, esp. 33 notes 46-49.
55
century BC.35 Recently, however, the view that the technique
was probably known in Greece earlier, even if used for other
purposes, has been supported. For example, the drainage
construction of the Minyans in Lake Kopais in the 15th century
BC is considered to essentially follow the qanat technique,36
while mythological references to the labyrinth of Daedalus at
Knossos are also connected with this particular water supply
system.37 In Greece, some scholars believe that we can talk
about the existence of water works based on the logic of qanat
(regardless of use) from the 6th century BC onwards. Some
of the cases are the 6th century BC pipeline in Athens,38 the
Tunnel of Eupalinos in Samos (540-525 BC)39 and the galleries
of Stryme (earlier than the 4th century BC),40 Olynthus (c. 420
BC),41 Perachora (c. 300 BC), Aegina, and Skyros,42 although
there is no written testimony that refers to the presence,
or the technique, of the qanat construction. Many relevant
constructions are dated to the Ottoman period. It is worth
noting that, already since antiquity, the qanat experienced
a rapid spread from the Middle East to Arabia, India, the
Iberian Peninsula and northwest Africa, especially in the
Roman period, while -in later tim and the
New World through the Spanish conquistadors (attested as
far as the city of Los Angeles).43
However, not all of the above-mentioned underground
aqueducts in Greece can certainly be called qanats. Today,
there are some reservations concerning the use of the
term qanat itself, and the wider term underground aqueduct
is preferred by some scholars. From a point of view, this is
not unjustified, since not all the underground aqueducts
are qanats. Nevertheless, the Hortiatis underground
aqueduct is indeed a qanat, and that is why we shall keep
using this term here for the specific construction.
The qanat of Haghia Paraskevi in Hortiatis (Figures 2-5) is the largest known in Greece.44 The water
abstraction zone45 covers an area of about 600m2, consists of underground tunnels (of a total length of 74m,
average height of 1.50-1.60m and width of 0.56-0.75m), three shafts46 and four water collection areas (Figure
2). The central part consists of three consecutive tun -SSE direction. In the SE part
The Persian origin of qanat is supported by the majority of scholars, such as Hodge 1992: 20; 2000: 35; Tölle-Kastenbein 1990: 39; Grewe 1998: 33. In
contrast, Forbes (1956: 665-666; 1964: 157-158) considers more likely that the qanat was invented in Armenia. However, in Iran, the qanat is still the
primary method of water supply, and it is estimated that in 1961 22,000 qanats were operating, with a total length of 170,000 miles, and a daily output of
48,384,000m3, equivalent of a large river (Wulf 1968: 94, 105; Hodge 2000: 37 note 4). Indeed, it is estimated that in Persia, in antiquity, there were about
40-50,000 qanats operating simultaneously, having a total length larger than the distance from the earth to the moon (Garbrecht 1988: 18).
Knauss et al. 1984:186 ff. Cf. Strabo 9.2.40.
Vavliakis 1997: 585-586; 2000: 47-48.
Perhaps of the Peisistratid period. See Thompson 1956: 49-52. Cf. Vavliakis 1997: 586; Grewe 1998: 55-56. A first description was given by
Curtius and Kaupert 1878:15,16.
Seitanidis 2003: 56 and 63, indicating, however, a difference between the Tunnel of Eupalinos and the qanat. Cf. Grewe 1998: 58-69. A bit later,
probably under Darius I, the Persians constructed the first qanat in Egypt, in the oasis of Kharga. See, e.g. Forbes 1964: 158. For the Kharga qanats,
which were probably constructed in the Roman period, according to recent research, see Schacht 2003.
Bakalakis 1967: 38-45.
Robinson 1946: 103 ff.
Cf. Vavliakis 1997, 586; 2000: 50.
Hodge 1992: 21-22; 2000: 35.
For the Haghia Paraskevi qanat, see Manoledakis 2007: 96 and in detail Manoledakis and Androudis 2007.
By ‘water abstraction zone’ we mean the section with the tunnels. Often, the term qanat is used to define only this part, although qanat can
be considered the entire construction, including the underground pipes that follow.
These shafts are cut for the cleaning and ventilation of the qanat while it is being constructed, and its maintenance afterwards.
Figure 2. Plan of the water abstraction zone of
the Haghia Paraskevi qanat.
Great Waterworks in Roman Greece
56
there is the largest water collection area, a rectangular arched room, from which two tunnels diverge to
different directions. One of them (E) is going SW, and ends ar the cave with the spring.
The other (is located at the lower level, accessible by a masonry scale and has a shaft at
each of its two ends (1 and 2). The depth of shaft 2 (7.90m) corresponds also to the maximum depth of the
whole qanat. From the NW end of this tunnel, as well as from the corresponding end of tunnel A
water continues its route in underground closed rectangular cross-section pipes, to its final destination.47
The tunnels are constructed in Quaternary sediments. The rocks in the qanat construction area consist
mainly of crystallized limestones of the Triadic Period and clay states of the48 The karst
origin of the water is indicated by stalactites 2-3cm long on the roof of tunnel E. The average water
production of the qanat is 22m3/h.49
The Haghia Paraskevi qanat is not just excavated in the soil, but has also masonry in many places,
consisting of bricks and semi-chipped stones. In the lower parts, bricklayers with crushed tile mortar
domain. Higher, we find stones with lime mortar, while in some parts there are remnants of brick masonry.
The entrance to the tunnels is located at the ground surface, at the north end of room , while the water
comes from the spring near the area .50
Construction phases and chronology
The dating of the construction of the Haghia Paraskevi qanat remains a difficult issue, as it applies to
every qanat in the absence of written or other evidence. According to the only view published so far, it
is considered as having been constructed in the Ottoman period, under Sultan Murat II.51 However, this
view clearly needs revision. It is mainly the preserved masonry parts in the qanat tunnels that can lead
to such a revision.
In the lower parts, where the water flows on the furrow (groove) walls, bricks are preserved at several
Manoledakis and Androudis 2007: 289.
The Haghia Paraskevi qanat has been geologically studied by Vavliakis et al. 1998: 73-87.
Vavliakis et al. 1998: 76, 81-87.
Manoledakis and Androudis 2007: 290.
Vavliakis 1997: 587; Vavliakis et al. 1998: 78. Indeed, Vavliakis, misinterpreting a passage from Tamiolakis’ book (Tamiolakis 2001: 18-20),
mentions that the qanat was constructed in the 18th century by Murat II, who, of course, lived in the 15th century.
Figure 3. Roman bricks on the
furrow walls of the Haghia
Paraskevi qanat.
57
4cm (Figure 3). In some parts lime mortar with grey-white and yellowish colour is preserved. Noteworthy
is the presence of a clay cylindrical pipe in tunnel A, 11cm thick and with a diameter of 26cm, of unknown
function, which could lead water from another nearby source to the groove, or perhaps serve the
ventilation of the tunnel. All these seem to belong to the first construction phase of the qanat, or at least
of its masonry. This phase certainly belongs to Late Antiquity. Indeed, it is likely to be Roman, judging by
certain elements: one is the similarity of the section of qanat to those of known Roman qanats in Europe,
such as the so-called Raschpëtzer-Qanat in Luxembourg, dated from the 2nd century AD.52 Another could
be the fact that the dimensions of the aforementioned bricks are absolutely identical to those of the
bricks in the lower parts of the buildings in the Galerius Palace in Thessaloniki.53 Essential for the dating
of the construction of the qanat in the Roman period is certainly the dating of the earliest part of the
Hortiatis aqueduct bridge in this period, as will be presented in detail below.
Based on the above, we can say that the Haghia Paraskevi qanat, which as a masonry construction existed
in Late Antiquity, may have been initially created during the Roman period. Thus, the view that there are
no known qanats in Greece constructed during the Roman and Byzantine periods must be refuted.54 This
first construction primarily may have been just excavated in the soil, with a simple masonry on the furrow
walls. Later, when the need to strengthen the structure, perhaps also to increase its production, appeared,
most of the parts of the brick masonry that we see today were created. Besides, the vaulted roofing of, for
instance, spaces and , with thinner and in many cases irregular bricks, presents similarities with the
Middle Byzantine period’s techniques (Figure 4). 55
Certainly, similar needs for improvement and expansion of the construction appeared also later, for
example, during the Ottoman period, and these works could indeed be connected to Murat II, for whom
we know that he cared for the repair and maintenance of the directly associated with the qanat Hortiatis
aqueduct.56 The extended parts that are built only with stones, such as the entire tunnel , seem to
belong to this period (Figure 5).57
Grewe 1998: 176 ff.
Manoledakis and Androudis 2007: 290.
Vavliakis 1997: 586.
Manoledakis and Androudis 2007: 291.
See infra.
Manoledakis and Androudis 2007: 291.
Figure 4. Part of Byzantine
masonry of the Haghia
Paraskevi qanat.
Great Waterworks in Roman Greece
58
Finally, we know that the last major repair of
the qanat took place in 1918, and this is visible
in several parts all over the monument, and
indeed it is often difficult to distinguish it
from the Ottoman ones. The restoration was
made by the Municipality of Thessaloniki,
to the ownership of which the whole water
supply system passed after the Liberation of
the city (1912) and until 1939.58 In 1975 the
qanat stopped watering Thessaloniki and
today it supplies with water a hospital, a
monastery and a military camp in the area
between Panorama and Asvestochori.59
The aqueduct bridge of Hortiatis
Description
The aqueduct bridge (today known as ‘the
aqueduct of Hortiatis’) is preserved in a total
length of approximately 223m and heads
northeast towards the village (Figures 1 and
6). To the opposite direction it continues
almost parallel to the Hortiatis-Asvestochori
road until it is lost under the territory of
a modern military camp. It practically
connects two hills, in order to keep the
required inclination of the water route to
the town steady.60
Tamiolakis 2001: 20, 57.
Manoledakis 2007: 96-98; Manoledakis and Androudis 2007: 291.
See in detail Manoledakis and Marki 2008.
Figure 5. Part of Tunnel A of the Haghia Paraskevi qanat,
constructed during the Ottoman period.
Figure 6. The south
face of the Hortiatis
aqueduct bridge.
59
The highest point of the
monument (499.07m above
sea level – 498.12m above
ground level) is to be found
at its southeast end, from
where the aqueduct heads
northwest for 118m (hereafter
Section 1, Figure 7). From that
point it changes direction and
presenting an angle of 140° it
heads west for another 101.5m,
until it meets the ground level
(hereafter Section 2, Figure 6).
The maximum height of the
aqueduct in the central part
of Section 2 reaches 20.1m,
and its maximum thickness
5m. In the centre of Section 2
the monument is divided into
three horizontal zones, from
which the lowest is slightly
thicker by about 1.00m than
the middle one and the latter
slightly thicker than the upper
one (Figure 6). On the upper
zone there is a water conduit of
rectangular section, measuring
0.5m across (Figure 8).The clay
water pipes were placed within,
however only few remains
of them have survived. The
conduit was covered with large
slates, many of which remain in
situ.61
In the centre of Section 2 there
are two large open arches
(Figure 6). The largest, in the
lower zone, has a height of
8.5m, a maximum width of
5.3m, and gives the impression
of an underpass on the dirt
road. Above this there is the
second open arch, with another
one to its west.
The north face of the aqueduct
(Section 2) is reinforced by
buttresses on either side of the large central arch, of which the two that are located to the east of the
arch (e.g. Figure 13, on the left) are preserved to a large extent – the western one, which is the largest, is
Manoledakis and Marki 2008: 361.
Figure 7. The southwest face of the aqueduct (Section 1) from the west.
Figure 8. Section 2. Slates covering the conduit on the aqueduct.
Great Waterworks in Roman Greece
60
wholly preserved – while to the west of the arch the remains of other buttresses are visible, damaged to
a large degree, so that it is difficult to determine their exact number.
On the south face, at a distance of 28.5m west of the western open arch, an arched room is opened into
the aqueduct masonry, which probably served the guard of aqueduct and its water (Figure 9). To the east
of the central arch there is a small brick-framed arch. At the western end of the aqueduct there is a water
collection area.
Section 1 is a straight stonewall, with a few rows of bricks only at its lower part (Figure 7). Only the
southwest face is actually visible, while the largest part of the northeast face rises just 0.5-1.5m from the
natural slope.62
The aqueduct is founded directly on the natural limestone rock. It is built with worked stones and
freestones of various sizes and bricks, while lime mortar has been used as a binder.63
Construction phases and chronology
The exact dating of the aqueduct, as well as the discrimination between its construction phases is not an
easy task. A careful observation leads to the conclusion that the monument received many repairs and
interventions throughout its history and that its image changed significantly several times. The view
that has been expressed and many times repeated in the past that in the three horizontal zones of the
aqueduct we can distinguish the three successive layers of its life phases, the Roman, the Byzantine and
the Ottoman ones (Figure 6), is simplistic and erroneous. 64 Much of what we see today is the product
of the Byzantine and the Ottoman periods. To the initial construction phase undoubtedly belong the
thick bricks (up to 8cm thick) in the lower part of the north face, which suggest that the aqueduct was
Manoledakis and Marki 2008: 361-362.
Manoledakis and Marki 2008: 363.
Tamiolakis 2001: 16, 24, who actually repeats the view of P. Theodoridis.
Figure 9. South side. Arched room at the western part.
61
initially built in the Roman period (Figure
10). It is likely that this construction was
part of a general activity of public buildings
construction, particularly water supply
ones, by the Romans during the Imperial
period. G. Bakalakis, reported in the mid-
20th century the incorporation of a Roman
altar into the aqueduct,65 which, however, is
nowhere visible today, nor was it identified
during the recent archaeological works on
the monument.66
based on the description from
O. Tafrali regarding a reservoir column
with a capital in the Citadel of Thessaloniki,
67 which received water from the Hortiatis
aqueduct, ventured a connection of this
reservoir to that of the Pergamon Acropolis,
which also had a column with capital; he
consequently ventured a dating of the
earliest water supply of Thessaloniki from
Hortiatis – and hence the construction of
the aqueduct- in the Hellenistic period,
during which, as he adds, the city had
already enough population.68 This view,
although interesting, is certainly difficult
to be proven, since it is based just on the
description of a structure that no longer
exists.
Until a few years ago, the dominant view on the dating of the monument was that of G. Tamiolakis – the
only one systematically involved with the aqueduct until the 2000s – namely that the monument was
built at the beginning of the 4th century, during the years of Galerius’ administration.69 In those years, as
he claimed, not only the necessary financial resources were available, but also many other indications of
the intention of the Romans to construct large water supply systems or bath facilities, in Greece generally
and in Thessaloniki particularly.
However, the thick bricks in the lower part of the north face, to the west of the large open arched, always
created to me the impression that they were earlier than the 4th century (Figure 10).70 Therefore,
within the frame of a study for the consolidation and conservation of the aqueduct in 2008, after the
initiative of the Municipality of Hortiatis and the former 9th Ephorate of Byzantine Antiquities,71 I sent,
after having taken the necessary license, parts of the aforementioned thick bricks to the Cultural and
Educational Technology Institute of the ‘Athena’ Research Centre in Xanthi, where they were dated with
the Thermoluminescence method. The dating AD 46 was given as the year of the firing of the bricks, with
an error deviation of about one century. Since dating becomes less likely as we move away from the year
Bakalakis 1953-1955: 361.
For this works see infra.
Tafrali 1913: 117.
Spieser 1984: 14.
Tamiolakis 2001:16. Cf. also Nigdelis 2006: 67.
Manoledakis 2007: 100.
This study (Marki et al. 2008) constituted practically the first archaeological and architectural study of the aqueduct in its long history. Its
results were presented in 2008 (Manoledakis and Marki 2008), while the results of the restoration works of the monument, which started in 2011,
based on the above-mentioned study, finished in 2014 and revealed some new data on the aqueduct, soon to be published by S. Akrivopoulou, the
archaeologist responsible for the works, and the author of this paper.
Figure 10. The thick bricks in the lower part of the north face. From
the initial construction phase of the aqueduct (Roman period).
Great Waterworks in Roman Greece
62
AD 46, we can generally say that the aqueduct must have been built in the 1st century AD.72 This makes
the dating of the first construction phase of the qanat in the Roman period even more likely, as was
mentioned above.
The history of the aqueduct is reflected in both its today’s masonry and in the written sources. The
monument we see today is characterized by the existence of different phases and repairs.73 It is
characteristic that we can distinguish traces of arches of the Roman and Early Christian periods that have
been later walled up (Figure 11). The today existing arches must belong to the Middle Byzantine period,
Manoledakis and Marki 2008: 364.
A detailed description of the monument and its phases, as they clearly appear after the recent archaeological research, will be soon released
by S. Akrivopoulou and M. Manoledakis (see above, note 71).
Figure 11. Parts of arches that have later been walled up, on the north side (Section 2)
of the aqueduct.
63
which means that the current image of the aqueduct significantly differs from the initial one.74 Some of
the construction phases that are distinguished can more easily be linked to specific historical periods and
events than others.
The surviving traces of arches indicate that in its initial phase the aqueduct probably had at least two
successive zones of close to each other arches, according to other examples of Roman aqueducts. Parts in
a lower zone on the south face, which clearly juts out from its overlying, later zones, belong to this initial,
Roman phase (Figure 12). In the western part of the south face a part of an arch is visible, which must
have been walled up no later than the mid-Byzantine period. Some relics of the same phase, with small
stones, thick bricks and parts of arches, are also preserved elsewhere.
A second construction phase, which appears to belong to the Early Christian period, is visible on the
north face, above the Roman part with the thick bricks to its west (Figure 13), as well as on the south
face, above and near the aforementioned Roman parts (Figure 12). Here, irregularly placed freestones are
interrupted by horizontal rows of bricks, slightly thinner than the Roman ones, while on the north face
several parts of arches are again visible, in the areas of the buttresses.75 This Early Christian phase could
be related to the water supply constructions that Constantine the Great is known to have carried out in
the area of Thessaloniki.76
The central part of Section 2, namely the part with the arches, belongs to the Middle Byzantine period
(Figures 12-13). It is the period during which the small arches of the earlier phases have been walled
up, the lower part of the monument has been reinforced with large stones, and the three arches we
see today have been opened, at least one of which received later interventions, judging by its pointed
Manoledakis and Marki 2008: 365.
Manoledakis and Marki 2008: 365.
Marki and Hatziioannidis 2004: 279-287.
Figure 12. The south face (Section 2) of the aqueduct. To the left of the large central arch, as well as in other places, parts of
arches are visible, which were later walled up.
Great Waterworks in Roman Greece
64
ending. Fieldstone and thin bricks dominate in the masonry. On the upper zone of the arches, to the east
of the central one and at the same distance from that as the western one, there was another arch, of the
same size as the western one, which gave an impression of symmetry. This arch was walled up with small
stones and thin bricks during the Late Byzantine (Palaiologan) period, when the small open arch that
exists today was created (Figure 12 and Figure 14). It is the only clearly visible Late Byzantine part of the
aqueduct.77
These Byzantine period repairs may be associated with the general development of the region of Hortiatis,
due to the building and the flourishing of the Monastery of Hortaitis.78 It is from this period that we have
the first indirect reference to the aqueduct from the Typikon for the Monastery of Christ Pantokrator of
1137;79 it is there implied that in the 12th century Thessaloniki was supplied with water from Hortiatis,
that the Monastery of Christ Pantokrator had rights over the income received from the water supply of
the city, and that the Hortiatis water was used in industrial activities.80 A few decades later, Archbishop
Eustathius of Thessaloniki, when describing the siege of the city by the Normans in 1185, refers to the
‘water from Hortiatis’ that watered Thessaloniki, to a leak in the reservoir of Eptapyrgion that the city
inhabitants suffered from, as well as to the efforts of the city’s governor to repair it.81 Similar references
are found in golden bulls of 1316 and 1317.82
Other parts of the monument are characterized by greater uncertainty as to their construction phase
(Figure 13, with stripes), which, however, dates from the Byzantine period and later. One or more
construction phases may belong to the Ottoman period, during which bricks are completely absent
Manoledakis and Marki 2008: 365-366.
For the Hortaitis Monastery, see just recently Manoledakis 2007:109-127, with the previous bibliography.
Gautier 1974: 121.
Cf. Letsas 1961: 146; Spieser 1984: 12; Manoledakis 2007: 102-104.
De Thessalonica capta, Bonn, 435-436 (S. Kyriakidis, Eustazio di Thessalonica, La espugnazione di Thessalonica, Palermo 1961: 78).
Actes de Chilandar I (ed. by M. Zivojinovic, V. Kravari, Ch. Giros), Paris 1998: 48, 61, 135-139, 230 n. 33, 6-10, 237 n. 34, 240 n. 35, 27-30; Dölger
1960: 66, 69; Manoledakis 2007: 104.
Figure 13. The north face (Section 2) of the aqueduct.
65
(Figure 13).The entire upper zone
of Section 2 (Figure 6), the arched
room of the south side (Figure 9),
and the large eastern buttresses
of the north side (Figure 13, on
the left), definitely belong to this
As regards Section 1, except for
a small zone in the lower part of
the southwest face showing the
features of the Early Christian
masonry, all the rest seems to
belong mainly to the Ottoman
period, perhaps with some traces
of earlier, Byzantine interventions,
confined on the southwest face
(Figure 7).83
During all these years, the
great significance of the water
of the Hortiatis aqueduct
for Thessaloniki is indicated
as undoubted through the
references of the written sources.
This significance becomes most
evident in a testimony delivered
by Hierax, the Grand Logothete of
Istanbul (16th c.), which refers to the era of the siege of Thessaloniki by the Turks in 1430.84 According to
this testimony, when Murat II was besieging the city, he encountered such a strong resistance that he was
ready to abandon his attempt. But then he was approached by the monks of the Vlatadon Monastery, who
advised him to cut off the water supply of the city from Hortiatis, so that the city residents would suffer
from water scarcity and surrender. The sultan, always according to Hierax, followed the advice of the
monks and put a soldier to guard the traitor monks, so that no one hurts them. According to a view, this
testimony of Hierax is based on a tradition, which corresponds to reality and explains the privileges of
the Vlatadon Monastery during the Turkish occupation, due to this very betrayal.85 However, the majority
of scholars consider the accusation of Hierax unfounded, even malicious, since no other relevant source
refers to such an event.86 It is indeed hard to believe that the defence of Thessaloniki was so strong in this
time to drive Murat to abandon the siege, 87 or that the reservoir of the Vlatadon Monastery was the only
source of water supply of Thessaloniki, so that its occupation would result in the surrender of the city.88
However, Murat II is definitely connected to the aqueduct of Hortiatis and specifically to the first of its
Ottoman phases, since shortly after the fall of Thessaloniki in 1430 he arranged a repair of the aqueduct
and built twenty new fountains in Thessaloniki, as well as the famous ‘Bey Hamam’, which received water
from it.89 Moreover, for the storage, cleaning and maintenance of the repaired aqueduct, Murat settled
some landless peasants in the area of the modern village of Asvestochori, in return for certain privileges.
Manoledakis and Marki 2008: 366.
Sathas 1872:256-257. Cf. also Tafrali 1913: 116; Spieser 1984: 12.
See the relevant views in Tamiolakis 2001: 27.
Dimitriadis 1983: 423; Letsas 1961: 147-148; Tamiolakis 2001: 26-27; Vakalopoulos 1969: 92; 1997, 194.
See e.g. Manoledakis 2007: 92-95.
Cf. also Tamiolakis 2001: 27.
Vasdravelis 1952: 167, text 129.
Figure 14. South face. Small open arch to the east of the central arches.
Great Waterworks in Roman Greece
66
Thus Yeni Köy (i.e. New Village) was founded, which, when its inhabitants began to produce and sell lime
in the 17th century, was renamed as Kireç Köy (i.e. Asvestochori).90
The operation of the aqueduct is also testified in the 16th century, in the diary of Gabriele Cavazza,
secretary of a Venetian ambassador who passed from Thessaloniki going to Istanbul,91 as well in the 17th
century, in the Seyahatnâme of Evliya Çelebi (1667), which refers twice to the Thessaloniki water supply
facilities from Hortiatis.92
Specific events concerning the aqueduct are mentioned in the 18th century: apart from the destruction of
the reservoir of the Vlatadon Monastery by rioters in 1714, which was accompanied by attacks on monks,93
another event is mentioned in a firman of 10 September 1722: ‘Some residents of Thessaloniki opened
up wells on Mount Hortiatis and managed to find large quantities of water, which they introduced to the
Hortiatis aqueduct. Then they sold part of it to the poor and kept the rest. In this way, however, water was
shut off in most of the city’s fountains, so that foreigners and the poor suffered from water shortage’.94
In the 19th century Michael Hatzi Ioannou called the aqueduct of Hortiatis ‘the most noteworthy
aqueduct of the present city (Thessaloniki)’.95 In 1927, nine years after the renovation of the Haghia
Paraskevi qanat, the Municipality of Thessaloniki stopped the water supply of the Vlatadon Monastery
from Hortiatis and channelled the latter’s water to new reservoirs.96 Since 1945 the aqueduct bridge was
not in use, but the water continued passing through the same point, just in an underground pipeline.
Thessaloniki continued to be supplied with water from Hortiatis until 1975.97
The preserved part of the Hortiatis aqueduct bridge is also connected to the most tragic event in the
history of the mountain and of the village in particular, since in its area the events that led to the Hortiatis’
Holocaust of 2 September 1944 where the slaughter of 149 people took place.98
Conclusion
Thessaloniki experienced a population increase and an economic boom in the Roman period, testified
from both written sources and the appearance of the first large public buildings in the city, among them
bath complexes. Thus, the underground waters of the city were no longer sufficient to meet its increasing
needs in water. The water of Mt Hortiatis, located very close to Thessaloniki, was an ideal, but also an
expensive solution. Therefore, an aqueduct was constructed, which, using the qanat tapping technique,
as well as several bridges – where required because of the natural relief – carried water from the Haghia
Paraskevi region of Hortiatis to the city, covering a distance of about 20km.
The dating of the building material of the first construction phase of the Hortiatis aqueduct showed that
the monument was built in the 1st century AD. This dating perfectly fits the reference by the Roman
authors to Thessaloniki as ‘the metropolis of Macedonia’ (Strabo 7 frg. 21) and as ‘the largest city of
Macedonia’ (Lucian, Lucius or the Ass 46). It also perfectly fits with the building development of the city
that is archaeologically documented at that time. Therefore, the Hortiatis aqueduct can be included
among the Roman aqueducts of Greece.99
Dimitriadis 1983: 217, 424; Tamiolakis 2001: 20. For Neochorion, see Hatzi Ioannou 1880: 49. Cf. also Tsekos 1957; Manoledakis 2007:105-106.
Mertzios 1947: 125.
‘The most delicious water in Thessaloniki is the one coming from the sources of Hortiatis, sweet, clean and very digestive. With this water the khans, mosques,
baths and public stores are supplied, as well as the mansions of prominent citizens’ (Seyahatnâme, book 8, ch. ‘Water supply’)…… ‘The freshwater of Hortiatis
comes to Thessaloniki in underground pipes and is distributed to all baths, mosques and the city’s imarets’ (Seyahatnâme, book 8, ch. ‘On the sources and rivers
of Mount Hortatzis’). For the water supply of Thessaloniki with the water of Hortiatis in the 17th century see also Vakalopoulos 1997: 231-233.
Stogioglou 1971: 54. In the same year the Hortiatis aqueduct was also seen by P. Lucas (1719: 56).
Vasdravelis 1952: 167, text 129. Cf. Spieser 1984: 12. For these texts, see also Manoledakis 2007: 106-108.
Hatzi Ioannou 1880: 66.
Stogioglou 1971: 54; Tamiolakis 2001: 26, 57, who also mentions the renovation of the aqueduct in 1918.
Tamiolakis 2001: 23.
Manoledakis 2007: 167-168, with the previous bibliography. All the details of the event with relevant documents can be found in Valachas
and Theochari 2008.
For the Roman aqueducts in Greece see Lolos 1997: esp. 302-312.
67
This is certainly not the case with the other two aqueducts that supplied the city with water, the aqueducts
of Retziki and Lembet, as the data available on these (considerably less than on the aqueduct of Hortiatis)
begin from the Byzantine period onwards. In this period the two aqueducts supplied the western part
of the city with water, while the aqueduct of Hortiatis watered the eastern part. In this case, it is highly
probable that the Roman fountains, baths and nymphaea of Thessaloniki, in both its eastern and western
parts, such those identified in the Dioiketerion Square,100 in the Agora,101 under the churches of Haghios
Demetrios, Panayia Acheiropoietos and Haghia Sofia, on Olympou, Phillipou, Profitou Ilia and Galileou
Streets, at the junctions of Egnatia and Antigonidon Str., Baltadorou and Venizelou Str., Armenopoulou
and Apostolou Pavlou Str., on Platonos, Sofokleous, and Haghiou Nikolaou Streets,102 in the Galerian
palace,103 as well as the ‘hydreion’ in the sanctuary of Serapis104, received water exclusively from the
aqueduct of Hortiatis, the only one confirmed in the Roman period.
In the same case, the inscription of the 2nd century referring to the water from the Altia region105 is also
likely to be associated with the Hortiatis’ aqueduct and this could mean the location of the settlement
called Altos somewhere around the route of this aqueduct. This is of course completely hypothetical,
and provided that all proposals made both to complete the inscription and to date the three aqueducts
are correct, and that Altos had not a different, yet unknown, aqueduct. Similar reservations about a
connection with Hortiatis exist also for the Roman and Early Christian water pipes found in the suburbs
of Thessaloniki, such as Thermi and Toumba.
Through the examination of the preserved parts of the aqueduct of Hortiatis, both in the qanat of Haghia
Paraskevi and in the aqueduct bridge at the entrance of the village, as well as through the references
of the texts from the 12th until the 20th century, it becomes clear that the aqueduct was an extremely
critical for the city of Thessaloniki public work, and for this reason it was repaired or even reconstructed
several times, during the Early Christian, Middle Byzantine, Late Byzantine and Ottoman periods.
Bibliography
Abbreviations used
BCH Bulletin de correspondance hellénique
IG X, 2 1 Inscriptiones Graecae, X: Inscriptiones Epiri, Macedoniae, Thraciae, Scythiae. Pars II, fasc.
1: Inscriptiones Thessalonicae et viciniae, ed. Charles Edson. Berlin 1972.
Ρωμαϊκή
Θεσσαλονίκη
To Archaeologiko Ergo stē Macedonia kai stē Thrakē (= AEMTh) 18: 223-238.
(ed.) Ρωμαϊκή Θεσσαλονίκη
To
Archaeologiko Ergo stMacedonia kai stē ThrakēAEMTh) 13: 191-206.
Makedonika 3: 353-362.
Voyage dans la Macédoine
Tassia 1993: 332-334.
Adam-Veleni 2004, who mentions that this fountain must have been constructed in the 2nd century BC, using the underground waters, which
the Agora area was rich in.
For these constructions, see Marki 1983: 14; Vitti 1996: 96, 174 nr. 45, 179 nr. 56, 191 nrs. 69 and 71, 194 nr. 81, 202 nr. 89, 204 nr. 90, 228 nr. 112,
229 nr. 115, 234 nr. 121, 239 nr. 126, 241, nrs. 127 and 129, 243 nr. 132. Cf. also the reservoirs and pipes in Vitti 1996: 192-193 nrs. 77-78, 197 nr. 82,
226 nr. 110, 243 nr. 130.
Vitti 1996: 110, 150-151; Athanasiou et al. 1999, with the previous bibliography.
IG X 2 1, 83. Cf. Vitti 1996: 175 nr. 47, with previous bibliography.
See above, in the first chapter.
Great Waterworks in Roman Greece
68
A Geographical and Historical Description of Ancient Greece vol. I. Oxford.
Atlas von Athen: 15.16. Berlin.
Τοπογραφία της Θεσσαλονίκης κατά την εποχή της τουρκοκρατίας 1430-1912. Thessaloniki:
Kyriakides Bros. Editions.
Dimitsas, M. 1874. Αρχαία Γεωγραφία της Μακεδονίας. Athens: Institute of Balkan Studies.
Regesten der Kaiserurkunden des Oströmischen Reiches von 565-1453, Teil 4 (1282-1341). Munich/
Berlin.
To Archaeologiko
Ergo stē Macedonia kai stē ThrakēAEMTh) 12: 217-222.
Archaiologikon Deltion 47: 424.
A History of Technology vol. II: 663-694. Oxford: Clarendon Press.
Studies in Ancient Technology vol. I. Leiden: Brill.
Garbrecht, G. 1988. Mensch und Wasser im Altertum. In: Frontinus-Gesellschaft e.V. (eds) Geschichte der
Wasserversorgung III, Die Wasserversorgung antiker Städte: 13-42. Mainz am Rhein.
Gautier, P. 1974. Le typikon du Christ Sauveur Pantocrator. Revue des études byzantines 32: 1-145.
Goblot, H. 1979. Les qanats: une technique d’acquisition de l’eau. Paris: Mouton Éditeur.
Grewe, K. 1998. Licht am Ende des Tunnels. Mainz am Rhein: Philipp von Zabern Verlag.
Hatzi Ioannou, M. 1880. Aστυγραφία Θεσσαλονίκης, ήτοι, τοπογραφική περιγραφή της Θεσσαλονίκης: περιγραφή
μιας εποχής. Thessaloniki: Makedonia editions.
Hodge, A. T. 1992. Roman Aqueducts and Water Supply. London: Duckworth Archaeology.
Hodge, A. T. 2000. Qanats. In: Ö. Wikander (ed.) Handbook of Ancient Water Technology: 35-38. Leiden: Brill.
Die Wasserbauten der Minyer in der Kopais – die älteste
Flussregulierung Europas
Thessalonikeon Polis 4: 217-224.
Letsas, A. M. 1961. Ιστορία της Θεσσαλονίκης 1961. Εταιρεία Μακεδονικών Σπουδών.Thessaloniki: Society for
Macedonian Studies.
Υδάτινες σχέσεις, Το νερό ως πηγή ζωής
κατά την αρχαιότητα: 39-46. Thessaloniki: University Studio Press.
Lolos, Y. A. 1997. The Hadrianic Aqueduct of Corinth (With an Appendix on the Roman Aqueducts in
Greece). Hesperia 66.2: 271-314.
Lucas, P. 1719. Troisième Voyage du sieur Paul Lucas, fait en M.DCCXIV par ordre de Louis XIV vol I. Rouen.
Archaiologikon Deltion 47: 424-425.
Manoledakis, M. 2007. Από τον Κισσό στον Χορτιάτη. Thessaloniki: Kornelia Sfakianiakis Editions.
To Archaeologiko Ergo stMacedonia kai stThrakAEMTh) 21: 285-292.
To Archaeologiko Ergo stē Macedonia kai
stē ThrakēAEMTh)22: 361-368.
Αρχαιολογία 7: 11-16.
To
Archaeologiko Ergo stē Macedonia kai stē ThrakēAEMTh) 18: 279-287.
Marki, E., Siaxabani, Ch., Manoledakis, M., Passia, A., Michailidis, G. and Vlysidou, Z. 2008. Study for the
consolidation and conservation of the Hortiatis aqueduct in Thessaloniki. Thessaloniki.
Mertzios, K. 1947. Μνημεία Μακεδονικής ιστορίας. Εταιρεία Μακεδονικών Σπουδών.Thessaloniki: Society for
Macedonian Studies.
To Archaeologiko Ergo stē Macedonia kai stē ThrakēAEMTh)5:
271-287.
Nigdelis, P.M. 2006. Eπιγραφικά Θεσσαλονίκεια. Thessaloniki: University Studio Press.
Makedonika 1: 377-383.
Byzantinische Zeitschrift 10: 23-39.
: πόλις και χώρα στην αρχαία Μακεδονία και Θράκη. Πρακτικά Αρχαιολογικού Συνεδρίου, Καβάλα
9-11 Μαϊου 1986
N. G. L.
Hammond
69
Papazoglou, F. 1988. Les villes de Macédoine a l’époque romaine. Bulletin de Correspondance Hellénique,
Suppl. XVI. Paris.
Robinson, M.D. 1946. Excavations at Olynthus. Part XII, Domestic and Public architecture
Sathas, K. N. 1872. Μιχαήλ Ατταλειάτης-Νικήτας Χωνιάτης-Θεόδωρος Μετοχίτης-Θεόδωρος Ποτάκιος-
Χρυσόβουλοι λόγοι-Ιέρακος Χρονικόν-Κατάλογοι των εν ταις Βιβλιοθήκαις του Αθώνος και του εν
Κωνσταντινουπόλει Μετοχίου του Παν. Τάφου χειρογράφων. Vol. II. Venice: Typois tou Chronou editions.
Mitteilungen des Deutschen Archäologischen Instituts, Abteilung Kairo 59: 411-423.
Seitanidis, G. C. 2003. Μορφολογικά-υδρολογικά και περιβαλλοντικά στοιχεία που συνδέονται με το Ευπαλίνειο
Όρυγμα της Σάμου. Thessaloniki.
Υδάτινες
σχέσεις, Το νερό ως πηγή ζωής κατά την αρχαιότητα: 17-37. Thessaloniki: University Studio Press.
Thessalonique et ses monuments du IVe au VIe siècle. Athens/Paris: Diffusion De Boccard.
Stogioglou, G. A. 1971. Η εν Θεσσαλονίκη πατριαρχική μονή των Βλατάδων. Thessaloniki: Patriarchal Institute
of Patristic Studies.
Topographie de Thessalonique. Paris: P. Geuthner.
Tamiolakis, G. 2001. Η ιστορία της ύδρευσης της Θεσσαλονίκης. Thessaloniki: University Studio Press.
To Archaeologiko Ergo
stē Macedonia kai stē ThrakēAEMTh) 7: 329-341.
Thompson, H. A. 1956. Activities in the Athenian Agora. Hesperia 25: 46-68.
Tölle-Kastenbein, R. 1990. Antike Wasserkultur
Tsekos, Ch. 1957. Ιστορία του Ασβεστοχωρίου. Thessaloniki (own edition).
Vakalopoulos, A. 1969. Ιστορία της Μακεδονίας 1354-1833. Thessaloniki: Vanias Editions.
Ιστορία της Θεσσαλονίκης 316 π.Χ.-1983. Thessaloniki: Kyriakides Bros. Editions.
Ιστορικά Αρχεία Μακεδονίας. Α’ Αρχείον Θεσσαλονίκης 1695-1912
Vavliakis, E. 1989. Τα συστήματα qanat στην Ελλάδα. Μελέτη συστημάτων qanat στην Επαρχία Φυλλίδας Σερρών,
από μορφολογική, υδρογραφική και κοινωνικο-οικονομική άποψη
Πρακτικά 1ου Διεθνούς Συνεδρίου για
την Αρχαία Ελληνική Τεχνολογία: 583-591. Thessaloniki: Piraeus Bank Group Cultural Foundation.
Υδάτινες σχέσεις, Το νερό ως πηγή ζωής κατά την αρχαιότητα
1: 47-50. Thessaloniki: University Studio Press.
Valachas, Th. and Theochari, D. 2008. 2 Σεπτεμβρίου 1944, Χορτιάτης. Ό,τι απόμεινε από τη μέρα εκείνη, η
μνήμη είναι. Thessaloniki.
Πρακτικά 4ου Πανελληνίου Γεωγραφικού
Συνεδρίου της Ελληνικής Γεωγραφικής Εταιρείας, 12-14 Οκτωβρίου 1995: 73-87. Athens: Greek Geographical
Society.
Vitti, M. 1996. Η πολεοδομική εξέλιξη της Θεσσαλονίκης από την ίδρυσή της έως τον Γαλέριο. Athens: En
Athinais Archaeologike Etaireia.
Wulf, H. E. 1968. The Qanats of Iran. Scientific American: 94-105.
Dr Manolis Manoledakis
International Hellenic University,
Thessaloniki, Greece
m.manoledakis@ihu.edu.gr
