The Auloi of Pydna
Stelios Psaroudakes
ZUSAMMENFASSUNG
Der Doppelaulos von Pydna wurde als einziges
Instrument seiner Art bei der dokumentierten Ausgrabung
einer ungestörten Begräbnisstätte gefunden.
Er ist in einem so gut erhaltenen Zustand, dass zu seiner
Vollständigkeit nur die Rohrblätter fehlen. Alle
erhaltenen Doppelauloi sind bisher ohne Provenienz
und ohne Kontext überliefert, wir wissen nicht einmal,
ob sie überhaupt paarweise aufgefunden wurden
(einen vergleichbaren Fund bildet nur das ausgegrabene
Rohrpaar von Akanthos, wobei dieses, was
den Erhaltungszustand betrifft, nicht vergleichbar
ist). Das Rohrpaar von Pydna hat eine kurze und
eine lange knöcherne Pfeife, die jeweils aus fünf Einzelteilen
besteht: Becher am Anblasende, Birne, Verlängerung,
daran ein Mittelteil mit vier Grifflöchern,
ein Ende mit zwei Löchern, das unterste jeweils ein
Windloch. Allem Anschein nach hielt der Spieler das
längere Rohr, mit dem Daumenloch mit Drehung im
Uhrzeigersinn, in der linken Hand und brachte darauf
den tiefsten Ton der kombinierten Skala hervor.
Auf dem kurzen Rohr, mit Daumenlochdrehung
entgegengesetzt des Uhrzeigersinns, war der höchste
Ton zu erzielen. Die jeweils vier anderen Klänge
waren die gleichen, die Löcher sind in fast genau
übereinstimmender Position angebracht.
1 INTRODUCTION
What has been established so far in our knowledge
of the ancient Hellenic double aulos, from an
organological point of view, can be summarized as
follows:
• The aulos comprises a pair of cylindrical pipes
having circular, more or less, finger holes, and
cylindrical bores throughout that are either in
one piece when made of wood1, or in several
sections when made of bone2.
• The pipes are operated by cane reeds, most
probably of the double kind, although the single
kind has not as yet been excluded from the
argument3.
• The pipes are not bound together in any way
but are physically independent of each other,
and their reeds are placed together inside the
player’s mouth, each hand of the player operating
on the holes of only one pipe4.
These are the basic established facts. However,
there are important questions, which have not as
yet been satisfactorily answered, and which have
to do both with the comparative physical aspects
of the pipes in the pair, and the comparative
acoustic behaviour of the two pipes played in conjunction
with each other.
From a physical point of view the pertinent questions
are:
• Are the two pipes equal or unequal in overall
length? Several depictions of the instrument
while being played suggest unequal lengths: the
left pipe is drawn longer than the right one (e.g.
Fig. 20). In frontal view, the left pipe is defi-
1 Athens-Daphn˙ (Psaroudak˙s, publication in preparation),
Elgin (Schlesinger 1939, 411–420, Pl. 17), Louvre (Bélis
1984a), Egypt (Behn 1954, Pl. 27; Psaroudak˙s 1994 Vol. 1,
271 with Vol. 2, Fig. 94). The Ephesos pipe (Hogarth 1908,
194, Pl. 37.12; Psaroudak˙s 1994, Vol. 1, 287, Vol. 2, 106) is
the only aulos made out of one piece of bone. It is, however,
much too small, possibly a child’s instrument.
2 Athens-Agora (Boulter 1953; Landels 1964) and Athens-
Akropolis (Psaroudak˙s 1994 Vol. 1, 261 with Vol. 2, Figs.
84–85; Psaroudak˙s 2002, 357, Pl. 19), Akanthos
(Κψαελης 1992, 93–95; Psaroudak˙s, this paper), Aphaia
(Furtwängler 1906, 429, Fig. 337), Argithea (Psaroudak˙s
2002), Argos (Psaroudak˙s 1994, Vol. 1, 267, Vol. 2,
Fig. 89), Brauron (Landels 1963), D˙los (Deonna 1938;
Bovon 1970, 233, Pl. 38), Delphoi (Psaroudak˙s 2002, 362,
Pl. 21), København (Olsen 1968), Korinthos (Broneer
1947, 241, Fig. 21, Pl. 61; Davidson 1952; Psaroudak˙s
2002, 358–361, Pl. 20), Korykeion Antron (Bélis 1984b),
Lindos (Blinkenberg 1931), Lokroi Epizephyrioi (Orsi
1917), PerachΩra (Dunbabin 1962), Rhodos and Ialissos
(Psaroudak˙s 2002, 356, Pl. 18), Sparta (Dawkins 1929),
Taras (Psaroudak˙s 1994, Vol. 1, 313, Vol. 2, Figs. 128–129).
3 West 1992, 83. For a good assessment of the relevant evidence
see Landels 1999, 28–31.
4 See, e.g., bell crater New York 25.78.66 of ca. 420 B.C.
(West 1992, Pl. 16), and amphora London E 271 of ca.
440 B.C. (West 1992, Pl. 21).
The Auloi of Pydna
Stelios Psaroudak˙s
nitely longer than the right one in a painting a
century earlier than the previous representation
(Fig. 21)5.
• Are respective holes of the two pipes in a pair
equidistant from the mouth? In other words, are
the hole patterns the same and at the same place
along the two pipes (adjacent to each other)?
• Are the two reeds physically equal in all
respects? Theophrastos6 seems to imply reeds
of a different form for the two pipes.
• Is one pipe in a pair intended to be the ‘left’ and
the other the ‘right’? And if their lengths are
unequal, on which side should the longer/
shorter pipe be held? Some written evidence
and a large number of vase paintings seem to
point in that direction.
Of course, physical differences between the
pipes would cause different acoustic behaviors. If,
indeed, the overall lengths of the pipes were different,
and the reeds were not exactly the same, and,
on top of that, the hole patterns were different and
at different places along the pipes, what would
then be the combined acoustic effect? The next
question concerns what kind of technique should
have been employed by the player, in order to
affect the intended acoustic behavior?
Although answers to these questions have been
attempted, what makes our evidence shaky is the
fact that auloi found in excavations are fragmentary
to such an extent, that they do not afford the necessary
evidence in order for us to tackle problems of
the kind just mentioned. Even in the case of pairs of
pipes brought to light, because these are products of
illicit excavations, we cannot be sure that the two
pipes belong together, that they are actually members
of the same pair7: the Elgin pipes, said to have
been found in Attica, on the way from Peiraieus to
Eleusis, surfaced in the late 19th century in the
British Museum, together with all the other items of
the Elgin Collection8; the København pipes, said to
come from Athens, surfaced in the National Museum
of København in 1961, with their sections obviously
reassembled in the wrong way9; as to the
Louvre pipes, their multiplicity of holes have rendered
them suspect as forming a pair10. So, unless a
complete instrument is found in a proper, controlled
excavation, the questions posed above will
not receive satisfactory answers.
2 THE AULOS PAIR OF PYDNA
It is very fortunate, then, that such a pair of auloi
was brought to light in 1996, during excavations in
one of the cemeteries of Pydna (dated to the first
half of the 4th century B.C.)11. Pydna was a large
harbor city in the kingdom of the Makedonai, a
strategic point on the coast of Pieria, in the
Aegean, which flourished in the 5th century B.C.12
In the Northern Cemetery of the city, Grave
324 (Field 951) contained a full skeleton and an
aulos pair; no other grave offering was found
(Fig. 1). The excavator was the archaeologist Manthos
Besios, of the 16th Ephorate of Prehistoric
and Classical Antiquities. The orientation of the
skeleton was from East to West, with the skull to
the East. The skull lay sideways, towards the right
shoulder. The upper ends of the auloi were placed
near the mouth, and the pipes ran parallel over the
bones of the right upper arm and shoulder of the
skeleton. The left hand lay on the ground and ran
parallel to the body. Undoubtedly, the iron nails
found at different levels in the grave belonged to
the wooden coffin, inside which the deceased had
been placed, and which has now disintegrated.
The instrument is kept in the Archaeological
Museum of Thessalonik˙ (No. Pydna 100).
3 REFERENCES TO THE PYDNA
AULOI THUS FAR
The instrument was presented for the first time by
the archaeologist Ourania Banou at the ‘1st International
Conference on Ancient Hellenic Technol-
198 Stelios Psaroudak˙s
5 However, in a number of Sicilian terracottas (from Fontana
Calda, Selinunte, Mozia, Mussomeli, Kamarina, Morgantina,
Adrano) dated to the 6th–3rd centuries B.C., depicting
women playing the double aulos amongst other musical
instruments, it is the right hand pipe which is often shown
longer (Bellia 2005, 62, Fig. 11, cover Fig. Also, Bellia 2006,
cover Fig.). According to Bellia (private correspondence),
about 50 % of these aulos-playing statuettes have the right
hand pipe longer, while in others the pipes are shown as
being equal (e.g. Bellia 2005, 59, Fig. 2). The fact that statistics
so largely diverge from either of the positions, namely,
longer pipe in the right or left hand, allows us, I believe, to
disregard the evidence of this particular material. It is
unlikely, on the other hand, that both types of aulos existed
(in Sikelia or anywhere else in the Hellenic world), one
with the long pipe in the right hand and another with the
opposite combination.
6 Enquiry into plants 4.11.7, Hort 1990, Vol. 1, 372; Barker
1984, 189, Note 12; Hagel 2004, 382, Note 61.
7 So Landels 1999, 42.
8 Schlesinger 1939, 411–420, Pl. 17.
9 Olsen 1968.
10 Adherents of the pair theory are: Bélis 1984a; West 1992,
100; Hagel 2004, 379–380, 385. Landels (1999, 279, Notes
19, 30), however, puts forward a strong case for the two
pipes being single pipes (monauloi): the holes of the Louvre
pipes, particularly the thumbholes, “show signs of shaping”,
which “suggests that they did not have keywork”.
11 For an aerial view of the site of Pydna see Μπ σις/Παππ
n.d., 7. The date of the cemetery is given in Μπνυ 1997, 519.
12 For an archaeological map of Northern Pieria see
Μπ σις/Παππ n.d., 4.
ogy’ in Thessalonik˙ in 1997, and a preliminary
report appeared in the published Transactions13,
together with a photograph of the pipes in situ
(Fig. 1), and a photograph of the pipes after
retrieval from the grave14. The pipes were among
the items in the exhibition ‘Gifts of the Muses:
Music and Dance in ancient Hellas’, set up by the
Hellenic Ministry of Culture in Bruxelles in 2003.
A brief description and a photograph of the instrument
(Fig. 2) appeared in the pages of the Exhibition
Catalogue15. In the same year (2003), in a
report about the exhibition in the Hellenic popular
archaeological magazine Corpus, a picture of the
pipes was again published16.
4 PRESENT INVESTIGATION
The instrument was studied afresh by the present
writer in March 200617. Let it be said at this point
that, because the auloi were methodically excavated
and photographed in place by the archaeologist,
there is no doubt that the sections are in the right
order. Even if they were not in the order in which
they were when the pipes were being played, at
least they are the way they were when the pipes
were placed next to the body. In other words, we
can be sure that no rearrangement of the sections
took place after the pipes had been buried in the
ground. It is also highly probable that the sections
are in the right organological order, but this, of
course, has to be proven. It must also be said at the
outset that the two pipes must have formed a pair
in antiquity, not only because they were found
together and are of similar construction and finish,
but also because the thumb holes in the ‘central’
sections (Figs. 18–19) are displaced with respect to
the upper series of holes in opposite directions, as
we shall see later.
By just looking at them as exhibited in the
Museum (Fig. 2), the pipes are obviously unequal
in length and have corresponding holes at different
distances from the mouth end. A closer look
reveals that corresponding distances between holes
are also not the same in the two pipes. Thus, one is
left with the impression that this particular aulos,
at least, comprises unequal pipes of different holepatterns
and at different distances from the mouth
end. However, as will be shown below, a closer
examination of the individual sections will draw a
somewhat different picture of this aulos’ form. For
the sake of convenience, let us call the two pipes L
and S that is, ‘long’ and ‘short’, respectively.
As is always the case with earlier finds, so with
the present aulos, the two pipes comprise five sections
connected to each other by the spigot-andsocket
method (Fig. 3): first comes a conical ‘cup’
at the mouth end that received the reed; then the
‘bulb’ section, immediately after without any
holes, and with the characteristic external bulge;
then the ‘extension’, a piece of tube without any
holes then the ‘central’ section with the first four
holes, including the thumbhole underneath as the
second hole; and finally the ‘exit’ section, with the
remaining two holes. The two pipes will here be
examined in parallel, one section type at a time.
4.1 THE ‘CUPS’ (SECTIONS TYPE ‘A’)
Cup aS (Fig. 4 top and Fig. 9 top left) survives in its
original length, thus providing us with the initial
length of the section (3.146 cm), and also with the
original depth of the socket receiving the reed
(socket depth 1.668 cm). Cup aL (Fig. 5 and Fig. 9
bottom left) has lost part of its entry length (surviving
length 2.580 cm). If we assume that the two cups
were equal in both effective18 length and socket
depth, then the missing length and the missing
depth of the socket of cup aL can be calculated
(missing length 0.566 cm; missing depth 0.604 cm).
Engraved into the outer surface of the spigot of cup
aS are about nine parallel lines (Fig. 4), most probably
for the purpose of joining together cup to bulb,
perhaps with the use of thread19. Similar indentations
were not, however, observed on the spigot of
cup aL (Fig. 5), but two parallel lines inscribed on its
periphery and near to the broken-off end were
clearly visible under the microscope.
4.2 THE ‘BULBS’ (SECTIONS TYPE ‘B’)
The bulbs are the sections adjacent to the cups
(Fig. 6 and Fig. 9 middle). They survive in their
13 Μπνυ 1997.
14 Μπνυ 1997, 522, Fig. 2.
15 Aνδρκυ et al. 2003, 177.
16 Anonymous 2003, 83.
17 I would like to thank wholeheartedly the former Director
of the Ephorate, Ms Polyxen˙ Adam-Velen˙, the former
Director of the Archaeological Museum of Thessalonik˙,
Mr D˙m˙trios Grammenos, and last but not least, the excavator,
Mr Manthos Besios, for, not only was I allowed to
examine and publish the find, but was also offered access to
the Excavation Diary and to the Restoration Laboratory of
the Thessalonik˙ Museum and its facilities. During my stay
in the Museum, I had the chance to discuss the find with
archaeologists Eleonora Melliou, Eutychia Kephalidou,
and Zo˙ Kotitsa, and restorer Sonia Athanasiadou. I would
like to thank them all, for sharing with me their knowledge,
inner thoughts, and enthusiasm.
18 By ‘effective’ length is here meant the length of tube seen
by the eye when the sections are joined together. Thus, the
spigot length is not included in the ‘effective’ length.
19 Landels (1999, 33) doubts the use of thread in joining the
sections together; he is inclined to think that the sections
were permanently glued together.
The Auloi of Pydna 199
original lengths (5.3 cm bulb bS, 5.82 cm bulb bL).
They are not exactly equal in length, which is
noticeable. It could be that as both bulb sections
were found in bits and glued together during
restoration, their original overall lengths have been
slightly altered, although the matches between the
bits seem correct. However, the fact remains that
the two sections differ now by just over half a centimetre
(0.52 cm), a fact that should not perhaps be
disregarded for the time being. The bulb sections
have sockets at their upstream ends, into which the
cups are inserted (Fig. 7).
Externally, very fine incised lines are clearly
seen on bulb bL (Fig. 6 bottom) at two places: two
together on the bulb side and one on its own on
the other, tubular, side. Similar lines can be seen on
bulb bS (Fig. 6 top), too, although not as clearly,
undoubtedly due to its eroded surface20. Here it is
proposed, rather hesitantly, that these lines, albeit
too fine to be distinguished from a distance, had a
decorative purpose21.
The maximum external diameter of bulb bL
(1.66 cm) is only just over a millimetre larger than
the external diameter of the remaining cylindrical
part (1.55 cm), and similarly for pipe S. This means
that the bulbs, at least in the case of the Pydna
pipes, cannot “have served to protect the reed
when the aulos was laid down on a flat surface”22.
They must have played a different role.
As with the Akanthos, Athens-Daphn˙ (most
probably), Berlin23, Elgin, København, Megara24,
Pompeii25, Reading26 and Taras pipes, and the
Pergamon model27, so the Pydna pair possesses
only one bulb per aulos. This now makes us think
that there never actually existed any auloi with
two bulbs per pipe, as some vase paintings seemed
to indicate28. Undoubtedly, what looks in iconography
like a “second bulb” near the mouth is simply
the combined curve of the stem of the reed and
the cup.
4.3 THE ‘EXTENSIONS’
(SECTIONS TYPE ‘C’)
Next in order come the extensions, cylindrical
pieces with sockets and spigots but without any
holes (Fig. 8 and Fig. 9 right). A close examination
of extension cS (Fig. 8 top) under the microscope
reveals a fracture at the entrance end: the remains
of a socket (surviving depth 0.33 cm) and an external
recession (surviving length 0.1 cm) are evident.
Obviously, part of the tube at entry is missing. The
question arises as to whether the missing length
can be established, and thus whether the overall
length of the section be calculated. The question
will be tackled after an examination of the corresponding
extension cL.
Extension cL (Fig. 8 bottom and Fig. 9 midright)
survives in all of its significant details, thus
its initial shape can be confidently restored. There
are spigots at both ends. The upstream spigot is
embraced by another section, a ‘ring’ or ‘band’ of
bone, 1.536 cm long (Fig. 8 and Fig. 9 bottom
right, rL). Part of the cylindrical surface of this ring
is missing, allowing us to see inside and to get a
good picture of the junction of the two sections.
The ring ‘sits’ on the upstream spigot, so that its
external surface is level with that of its host and
extends beyond the spigot by just under a centimetre
(0.948 cm), therefore, together with the
spigot underneath, forming a socket of that depth
(0.948 cm in depth). On its external surface, near
its entry, and all around its circumference, the ring
has a shallow depression (0.444 cm wide).
This geometric feature reminds us of similar
constructions in other known auloi: the Elgin
pipes (Fig. 10) and the Athens-Daphn˙ pipe (Fig.
11), both wooden, and in one piece; the Athens-
Agora G fragment (Fig. 12); the Lindos F fragment
(the only example with a hole I in it) (Fig. 13); the
PerachΩra G and H fragments (Fig. 14); the PerachΩra
I and J fragments (Fig. 15), all, undoubtedly,
extension sections. In these ‘troughs’, at the
points where the bulbs are inserted into the extensions,
metal or leather bands might have been
inserted (although no trace of any material is left
on any such section), in order to prevent the bone
from splitting but also securing a better grip of the
bulbous sections. Very importantly, the widths of
these troughs are equal to those of the two Elgin
pipes (Fig. 10), the only pair with this feature on it
– assuming, of course, that the two pipes do form a
pair29.
One could safely, I believe, take it from there
and generalize that in a pair of pipes these portions
20 Similar incised lines have been observed on the bone ‘bulb’
Fragment A from Athens-Agora (Landels 1964, 394). In
the same group of aulos fragments, incised lines appear on
Fragment B (Landels 1964, 395); Fragment E (Landels
1964, 398); Fragment H (Landels 1964, 400).
21 Landels (1999, 34) also speaks of bone sections turned in a
lathe and “decorated with incised lines”.
22 Landels (1999, 33), one of three suggestions made on the
function of the bulbs in auloi (the other two being decoration
and ballast).
23 Behn 1954, Pl. 27.
24 Psaroudak˙s 1994, Vol. 1, 301–302, Vol. 2, Fig. 118.
25 Behn 1954, Pl. 58.
26 Landels 1968, Pl. 5.
27 Behn 1954, Pl. 58.
28 Landels 1999, 27, 32, Figs. 2a.3, 2a.9.
29 West (1992, 100, Note 84) is sceptical about the Elgin pipes
being a pair; compare Landels (1999, 279, Note 30): “The
Elgin auloi may have been a pair […]”. However, the comparison
made between the Elgin pipes and the Pydna pipes
below will strongly, almost conclusively, support the pair
theory.
200 Stelios Psaroudak˙s
of the extensions were of the same geometry, at
least when seen from the outside. If so, and since
the bottom part of the socket and the beginning of
the depression on extension cS survive, the missing
length of extension cS can be calculated (missing
length x = 0.544 – 0.1 = 0.444 cm; Fig. 16). The ring,
therefore, is not really an independent member of
the pipe, but rather an ‘elongation’ of its extension
section. The effective lengths of the two extensions
are now known (3.50 + 0.544 = 4.044 cm for extension
cS, 4.792 + 1.536 = 6.328 cm for extension cL),
their difference being 2.284 cm, a significant
amount; this, then, is responsible for the non-correspondence
of the first holes (I) of the pipes.
It is not apparent why the aulos maker constructed
the two extensions in a different manner,
one with and another without a ring. Possibly,
extension cL already existed, and, in order to make
use of it rather than discard it, he thought it easier
and, perhaps, more economical to supplement it
with a ring rather than to make a whole new section
from scratch. Or, alternatively, the ring was
introduced at a later stage, during repair work: the
initial cL was broken and was replaced by another
cL section, which was readily available but not
long enough; thus a ring was made to supplement
it. Let it be said that this feature in unique to this
aulos and is not found on any other surviving
extensions.
4.4 THE ‘CENTRAL’ PARTS
(SECTIONS TYPE ‘D’)
These sections are cylindrical and have a set of
four holes in them (here called I, T, II, III): three
(I, II, III) on the same side and in line with each
other (Fig. 17), and one on its own (T) on the other
side (Fig. 18), not exactly diametrically opposite to
the set of three holes. Numerous such sections
have been reported in the relevant bibliography30.
Both section dS and section dL have a socket at the
upstream end and a spigot on the other side.
The sections are of unequal length: dS overall
12.660/effective 11.952 cm; dL overall 13.65 cm/
effective 12.856 cm31 (Fig. 19). The spigot of section
dS has obviously lost part of its material: not only
its edge seems to have worn off (surviving length
0.638 cm), but also it is shorter by about 2 mm than
the receiving socket of the next section (0.806 cm).
The difference in the effective lengths of the two
sections is just under a centimetre (0.904 cm), not an
insignificant amount. Obviously, the reason why
section dL is longer than dS is because more space
was needed in order to accommodate the larger distances
of the holes in section dL.
On both sections the upper holes (I) and their
respective thumb holes (T), as has already been
said, are not diametrically opposite (Figs. 18–19).
The thumb holes are shifted slightly to the side,
counter-clockwise (as one is looking down, along
the stream direction) in the case of section dS, and
clockwise in the case of section dL. Central sections
with displaced thumb holes are reported in
literature32, where the feature has been interpreted
as an indication of right or left pipe33. By trying
out fingering these sections, I am inclined to think
that a clockwise shift indicates a left hand pipe and
a counter-clockwise shift a right hand pipe. There
is good evidence for that, as will become apparent
immediately below:
• The evidence of iconography. Very interestingly,
in iconography showing auloi while being
played, the pipe held in the left hand is nearly
always drawn longer than the other one34. In
Paquette’s iconographic collection35 the longer
pipe is held in the left hand fifteen times (A4,
A3, A736, A8, A1037, A14, A16, A23, A27, A34,
A36, A37, A43, A52, A54) and only twice in
30 Athens-Agora C, D, E (Landels 1964, 393, Fig. 1 A, C, D);
Athens-Akropolis A, B (Psaroudak˙s 2002, 357, Pl. 19 A,
B; Argithea D (Psaroudak˙s 2002, 344, Pl. 7 D); Brauron
(Landels 1963, 116, Fig. 2, Section AB); København
(Psaroudak˙s 2002, 364, Pl. 24); Korinthos F (Psaroudak˙s
2002, 359, Pl. 20.2 F); Korinthos G (Psaroudak˙s 2002, 360,
Pl. 20.3 G); Lokroi Epizeuphyrioi (Psaroudak˙s 1994, Vol.
2, pocket, Fig. 116); PerachΩra T, U, V, X, I΄ (Psaroudak˙s
1994, Vol. 2, Pl. 121 T, U, V, X, I΄); Sparta D, E, F
(Psaroudak˙s 1994, Vol. 2, Fig. 126 D).
31 Landels (1999, 33) mentions that the sheep or deer tibia
bones were most suitable for making aulos sections, and
that these were usually not longer than 9 cm. However, the
central and exit sections here of both pipes exceed this size.
Similarly, the København pair comprises central sections of
overall lengths 11.1 and 12.6 cm, and exit sections of 9.9
and 10.1 cm, all larger than 9 cm. Also, longer than 9 cm are
the sections: Athens-Akropolis A (14.5 cm), B (over 13 cm);
Argithea A, D, E (10.25, 10.78, 9.5 cm, respectively); Brauron
(13.1, 10.8 cm); Ephesos 13.4 cm; Korinthos C, F, I, H,
G, L (14.65, over 12.7, over 10, 10.6, 14.07, 12.2 cm respectively);
Korykeion Antron B (11.1 cm); Lokroi Epizeuphyrioi
(over 16.4, 13.6 cm); PerachΩra T, B΄, C΄, D΄, E΄, I΄
(over 12.3, 12.4, 11.1, 10.9, 10.3, 19.2 cm resp.); Sparta I
(9.37 cm) – see, collectively, Psaroudak˙s 1994.
32 Brauron (Landels 1963, 117–118); Elgin-Long clockwise T,
Elgin-Short anticlockwise T (Psaroudak˙s 1994, Vol. 2,
pocket, Fig. 105); Korinthos G+H clockwise T
(Psaroudak˙s 1994, Vol. 2, Fig. 113G+H); PerachΩra T
clockwise T (Psaroudak˙s 1994, Vol. 2, Fig. 121T);
PerachΩra U anticlockwise T (Psaroudak˙s 1994, Vol. 2,
Fig. 121U).
33 For example, in the case of the Brauron aulos, the clockwise
rotation of the thumb hole has been thought of as an
indication of a right hand pipe, although the elliptical
recessings on holes IV and V are in line with the fourth and
little fingers of the left hand: “two contradictory items of
evidence” (Landels 1963, 117–118).
34 For an exception to this ‘rule’ see No. 5, above (Sicilian terracottas).
35 Paquette 1984, 39–61, Pl. II; 21, Fig. 21B.
36 Reproduced here as Fig. 20.
37 Reproduced here as Fig. 21.
The Auloi of Pydna 201
the right hand38. It has been suggested that the
apparent difference in the lengths of the two
pipes in iconography should be interpreted as
the painter’s attempt to show depth of field:
“One [pipe] may appear to project further than
the other, but this is almost invariably the
artist’s way of trying to indicate that it is nearer
to the viewer”39. This is true, of course, in the
case of auletes facing to the left, whence the
left, longer pipe is closer to the viewer40, but
does not hold true in the cases of pipers facing
to the right, in which case the pipe closer to the
viewer, the right one, is drawn shorter41. There
is no doubt that the painters knew that the
pipes were unequal42, and in most cases they
tried to indicate this feature in their work.
• The evidence of the Akanthos pair. The Pydna
pipes are comparable to another pair of auloi,
found some years earlier (1988) in a grave (Burial
4769) in one of the cemeteries of Akanthos
(Field 165), another Macedonian town in
Chalkidik˙ on the Aegean coast43. These pipes,
alas not as well preserved, afford strong evidence
of the argument presented here, namely,
that a clockwise thumb hole belongs to a left
hand pipe and vice-versa: there are marks like
‘eyebrows’ on the mouth end side of the thumb
holes, cut into the bone and orientated in
opposite directions to each other, which, no
doubt, indicate whether the respective holes
were operated by the left or the right thumb
(Fig. 22). If one places one’s fingers on the sections,
one discovers that the hole with the
clockwise shift and the ‘acute’ mark fits the left
hand (Fig. 23), while the hole with the counterclockwise
shift and the ‘grave’ mark fits the
right hand (Fig. 24): marks and thumbs have
the same direction.
Furthermore, the section with the ‘acute’ mark is
the longer of the two Akanthos ‘central’ sections
(Fig. 25), indicating that, by analogy with Pydna
section dL, it belonged to the longer pipe of the
pair. Unfortunately, the Akanthos pipes do not
survive in their entirety, so the argument that the
longer ‘central’ section belongs to the longer pipe
of the pair cannot be proved for this particular
instrument. However, as with the Pydna pair, so
with the Akanthos pair, the exit sections are not
exactly equal in length (10.494 cm and 9.880 cm).
Of the two exit sections it is the longer one that is
connected to the longer ‘central’ section, indicating
that the longer sections (d and e) belong to the
longer pipe in the pair (Fig. 25). Thus it can, with
little hesitation, be suggested that, in general:
longer sections belong to the longer pipe in a pair,
which is held in the left hand.
In the case of the Elgin pipes, too, the thumb
hole with the clockwise shift appears on the longer
pipe, an indication that this was the left pipe
(Fig. 26). Interestingly, the longer Elgin pipe is of a
lower register than the short one. This is in accord
with the Pydna pair, where the lower register pipe
is the left hand pipe. Thus, the above observation
can now become a general rule (the ‘4L Rule’):
Longer sections belong to the Longer pipe
in a pair,
which is held in the Left hand and is of
Lower register44.
All four holes on both central sections of the
Pydna aulos, dS and dL, have a ‘dip’ around the
periphery of the holes, that is, the edges of the
holes are not ‘square’ but have been gently
smoothed out (Figs. 17–18). This is especially evident
in section dS around holes III and T, less so
around I and II, as these latter holes are not in a
good state of preservation, due to corrosion of the
surface around them. On section dL all four holes
have the recession. Similar recessions have been
observed on many extant aulos fragments and have
been thought of as having provided a better ‘seating’
for the fingers45, for better air sealing.
38 Paquette 1984, 21, Pl. II, Fig. B; 53, Fig. A38.
39 Landels 1999, 42.
40 E.g. Paquette 1984, 39, 41, 43, 47, 51, Figs. A5, A7, A14,
A16, A25, A27, A34.
41 E.g. Paquette 1984, 39, 40, 45, 47, 53, 57, 59, Figs. A3, A4,
A6, A8, A23, A24, A37, A52, A54.
42 See, especially, Paquette (1984, 49, Fig. A31), where the
two, definitely unequal pipes are clasped in one hand by
the musician, and there is no question of perspective.
43 I would like to thank the Director of the 16th Ephorate of
Prehistoric and Classical Antiquities, Thessalonik˙, Ms Lilian
Acheilara, for granting me permission to examine the
Akanthos auloi, and the archaeologist in charge of the
excavation, Ms Helen˙ Trakosopoulou, for consenting to
it. I am indebted to Mr Micha˙l Larentzak˙s-Laskaris,
restorer at the Ephorate, for his friendly reception and our
illuminating discussions during my weekly visit to the Laboratories
of the Ephorate. Special thanks are also due to Ms
Athanasia Raïs˙ and Svetlana Vivtenko, restorers, for their
help with microscopic examination of certain details and
their good will to share with me their knowledge on material
decay. The aulos pair is kept today in the Archaeological
Museum of Polygyros, Chalkidik˙ (Inventory Number I
165.28 = Ierisos, Field 165, Item 28). A preliminary report
on the instrument, with photograph, is given in
Κψαελης 1992, 93–95.
44 In the case of the Louvre pipes, Bélis (1984a, 113) reports
that only pipe ‘A’ [the ‘high register’ pipe] exhibits a thumb
hole shift, although the direction of this shift is not specified.
Of pipe ‘B’, [‘low register’], it is said that all holes are
absolutely aligned (“ici, les trous sont rigoureusement
alignés”). The absence of opposite shifts of the thumb holes
in the Louvre pipes may now be thought of as further, perhaps
not very strong, evidence against the argument of
them forming a pair. See footnote 10, above.
45 Landels 1999, 34.
202 Stelios Psaroudak˙s
4.5 THE ‘EXIT’ SECTIONS
(SECTIONS TYPE ‘E’)
The last sections of the pipes are cylindrical,
unequal in length, and have in them two holes each
(IV, V*)46 on the same side (Fig. 27). There are no
holes on the lower side (Fig. 28). At the upstream
end, both sections, eS and eL, have a socket (Fig. 29).
As regards finger ‘seating’, this can be found on
holes IV, but it is certainly absent from holes V*
(Fig. 27). Similar absence of ‘seating’ around the last
hole of several aulos exit fragments is reported in
literature, and it has been interpreted as an indication
that these holes were not fingered; they were
mere vent-holes47. Indeed, as each hand of the aulos
player has only five fingers, with the thumb fixed in
place, a sixth hole cannot be handled without the
use of some remote control device.
4.6 ASSEMBLING THE PARTS
When the individual sections as they survive today
are joined together in the order cup-bulb-extension-
central part-exit part (a-b-c-d-e), two pipes
are formed of unequal length (S, L) and non-corresponding
holes (Fig. 2). With the proposed reconstruction
of the extension cS the difference in
length of the two pipes increases, and a much better
correspondence of the tone holes is afforded:
correspondence of upper holes (Fig. 30); correspondence
(approximate) of thumb hole TL with
hole IIS (Fig. 31); correspondence of thumb hole
TS with hole IL (Fig. 32).
So, there seems to be a logic behind the placing
of the hole patterns in the two pipes, as shown
schematically below48:
46 The asterisk indicates a vent-hole.
47 Landels 1999, 34: “The vent-hole […] can be distinguished
by its sharp outer edge, not shaped for the player’s fingers”.
48 The diacritic → means that the hole which bears it is located
a little further down from its counterpart (e.g. TL → is a
little lower than IIS). Absence of the diacritic indicates
good correspondence of holes.
49 It must be remembered that the long pipe in the Elgin
pair is quite deformed, being exceedingly curved, which
makes it difficult to take reliable measurements; compare
Landels (1999, 279, Note 30): “They [the Elgin auloi]
The Auloi of Pydna 203
L I T→ II→ III IV V*
S I T II III IV V*
This correspondence pattern is very much like that of the Elgin pair: correspondence of upper holes
(Fig. 33); correspondence (approximate) of thumb hole TL with hole IIS (Fig. 34); correspondence (approximate)
of thumb hole TS with hole IL (Fig. 35), as shown schematically below49:
L I→ T→ II→ III→ IV→ V*
S I T II III IV V*
Interestingly, the difference in the lengths of the pipes at these hole correspondences are: 3.138 cm for the
Pydna pair and 3.2 cm for the Elgin pair50 (Fig. 36), practically the same length difference. The effective
lengths of the Pydna sections are tabulated below, together with the overall effective lengths of the pipes
and the difference between them (in cm):
• (Μπνυ 199751)
a b c d e overall effective length pipe difference
L … 6.7 … 6.4 12.9 11 37.0 2.8
S … 8.0 … 3.5 11.7 11 34.2
• (Psaroudak˙s52)
L <2.256> 4.872 6.328 12.856 11.040 37.352 3.138
S 2.256 <4.872> 4.044 11.952 11.090 34.214
The most noticeable difference between the figures in the above table is that in the sizes of section cL:
3.5 cm (surviving) and 4.044 cm (restored). This is a very important point, for the extra 0.544 cm added to
have become so badly distorted that measurements are
unreliable”.
50 The Elgin pipes are 31.2 and 34.4 cm long (excluding the
problematic bulb sections); see Psaroudak˙s (1994, Vol. 1,
282), where it is argued that the bulbs are wrongly placed
on the pipes.
51 Μπνυ (1997) does not give measurements for cups and
bulbs separately.
52 The diacritic <…> indicates that the enclosed figures are
derived from the other pipe of the pair, assuming equality
of the corresponding sections.
53 The first figure gives the diameter reading along the axis of
the pipe, and the second that at right angles to it.
54 By XL–XS is marked the difference between the distances
along the two pipes of corresponding holes from the cup
ends to their centres.
55 Hagel (2004, 380–385, Diagrams 1–4, Tab. 1–3) reports to
have developed a computer programme, which calculates
the pitches of the notes produced by the holes of any
metal-wrapped aulos pair (i.e. with rotating metallic
‘sleeves’ over the holes, 2.5 mm thick). The programme is
applied to the Louvre pipes – on the belief that they formed
a pair and were operated by rotating ‘sleeves’ – and the
results are presented and assessed. With mouthpiece extrusions
of certain values (4.23 cm for the ‘high’ pipe A, and
4.58 cm for the ‘low’ pipe B) the lowest notes of the pipes
were near identical (177.9 Hz for A, and 177.8 Hz for B),
while corresponding holes on the two pipes produced nearequal
notes, differing by as low as 0.5 Hz (couple Vh–IIl),
to as high as 7.6 Hz (couple IVh–Tl). Interestingly, it is the
‘h’ holes (pipe A) in the couples which are a little higher
than the ‘l’ ones (pipe B), apart from couple VIIIh–Vl,
where note VIIIh is lower than Vl (by 2.4 Hz). Note VIl of
pipe B (203.8 Hz) has no ‘h’ counterpart in pipe A. The
software is not presented in the publication, so it cannot be
assessed here.
The table shows that the distances of the holes
from the cup end in the long pipe are in every case
a little larger than those of the corresponding holes
in the shorter pipe (XL–XS values). Although a
good match between the couple IL and TS is
afforded, TL and IIS show a maximun diversion,
which is then reduced in the remaining hole couples.
One could, therefore, say that, apart form the
couple IL–TS, the rest of the hole couples are
almost perfectly aligned. However, a different picture
is given by the Elgin pair, where there is a distinct
and uniform shift of the L and S sets of holes.
A set of questions, therefore, arises: is the shift
observed in the Elgin pipes deliberate, or is it due
to the distortion of the pipes, especially the long
one? Are the Pydna pipes closer to reality as
regards the respective placement of the holes? In
other words, should we think of the L and S sets of
holes as corresponding, more or less, with each
other, or are we to allow for a deliberate displacement
of the two sets at the stage of manufacture?
In the case of the Pydna pipes, does the close correspondence
of the holes in the two pipes suggest
(near or exact) identity of notes produced by these
holes? Obviously, answers to these questions can
only be given after an experimental study carried
out on a replica of the instrument, in an attempt to
establish notes and scales, and investigate possible
heterophonic use of the pipes55.
5 CONCLUSIONS
• The short Pydna aulos was longer in antiquity
than what it is today by about half a centimetre
(0.444 cm).
• The two pipes of the Pydna pair were unequal
in length, differing by 3.138 cm.
• The longer sections belonged to the long pipe.
• The short pipe was held in the right hand and
the long pipe in the left hand.
• The left, longer, pipe was of a lower register
than the right, shorter, one.
• The highest note of the pair was produced by
(opening) the top hole of the short pipe alone
(IS), while the lowest note of the pair was produced
by (closing) the lowest tone hole of the
long pipe alone (IVL).
204 Stelios Psaroudak˙s
pipe S brings its holes in better alignment with those of pipe L (Fig. 30). The diameters of the holes are as
follows (in cm)53:
I T II III IV V*
L 0.9x0.9 0.884x0.846 0.854x0.894 0.876x0.856 0.848x0.886 0.912x0.872
S 0.746x0.84 0.922x0.922 0.862x0.894 0.894x0.880 0.804x0.90 0.824x0.912
The distances of the (centres of the) holes from the mouth end are tabulated below (in cm):
cup I T II III IV V* exit
L 0 16.456 19.576 22.189 25.204 28.502 33.23 37.352
S 0 13.879 16.391 18.961 21.983 24.872 28.236 34.214
The degree of correspondence of the holes between the two pipes can be discerned from the following
table54:
16.456 19.576 22.189 25.204 28.502 33.23 37.352
L – IL TL IIL IIIL IVL VL* exitL
S IS TS IIS IIIS IVS VS* – exitS
13.879 16.391 18.961 21.983 24.872 28.236 34.214
XL–XS 0.065 0.619 0.206 0.329 0.266
• The patterns of holes in the two pipes correspond
well between them in general, although
the long pipe set seems to be displaced a little
lower down with respect to the short pipe set.
• The same kind of correspondence (vii) seems to
exist between the sets of holes in the two Elgin
pipes, although there the set of holes of the
long pipe is clearly lower than that of the short
pipe (unless, of course, the discrepancy has
been caused by an excessive distortion of the
wood).
• Although the bulbous sections of the Elgin
pair are not in a good state of preservation, it
is evident from the rest of the pipes that the
Pydna pair and the Elgin pair were instruments
of the same kind, but perhaps of a different
pitch level, the Pydna aulos being of
lower register.
The Auloi of Pydna 205
