Kadmos 2022; 61(1/2): 61–96
Athanasia Kanta, Thomas G. Palaima, Massimo Perna*
A Hieroglyphic seal from the cult centre of
the city of Knossos (KN S (4/4) 01)
with an appendix by Alessandra GiumliaMair
https://doi.org/10.1515/kadmos20220003
Abstract: The Hieroglyphic seal KN S (4/4) 01, found in the Neopalatial Room 3 of
the Cult Centre of the City of Knossos, is a document of considerable importance
in the study of Cretan script for several reasons.
With regard to iconography and palaeography, we have identified for the first
time the true idealized form of sign 095, and that permits us to correct our understanding of the shape of other examples of the same sign and thereby to improve
our knowledge of the palaeographical development of the Hieroglyphic script.
Furthermore, this new document preserves one more attestation of the socalled Arkhanes “formula” and shows a connection between this “formula” and
the fraction signs.
Finally, this seal confirms also a relationship between this “formula” and the
logogram *181 on the one hand and the sign 010 on the other hand.
Keywords: Hieroglyphic seal – Arkhanes “formula” – Hieroglyphic sign 095.
Article note: This excavation and the study of the material, which still continues, would not
have been possible without the very generous backing of the Institute for Aegean Prehistory.
We are very grateful to Dr. Malcolm Wiener, founder of INSTAP, Professor Philip Betancourt,
Executive Director of INSTAP, and Dr. Tom Brogan, director of the INSTAP Study Center in Crete,
for their unfailing and continuous support during the years of the work at Knossos. We are
also grateful to the director of the Heraklion Ephorate Dr. Vasso Sythiakaki, and section leader
Ioanna Serpetsidaki, for their help and support. We also thank the conservators of the Hera
klion Ephorate, Georgia Pachaki, Manolis Hatzimanolis and Elpida Politaki, the guard Stavros
Amanakis, and finally the excavation team of our archaeologists and technicians.
*Corresponding author: Massimo Perna, Centro Internazionale per la Ricerca sulle Civiltà Egee
“Pierre Carlier”, Portico Giovanni Corrias, Oristano – 09170, Italia.
EMail: maxperna59@gmail.com
Athanasia Kanta, Odos Myronos Nikolaidi 6, 71305, Iraklio, Greece.
EMail: athanasiaka@gmail.com
Thomas G. Palaima, Department of Classics WAG 123, University of Texas at Austin, 2210 Speed
way MD C3400, Austin, TX 787121738, USA. EMail: tgpalaima@gmail.com
Alessandra Giumlia-Mair, AGM Archeoanalisi, Merano BZ, Italia.
62
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Excavation and stratigraphy (A. K.)
In a plot at the modern west village of Knossos, now called Bougadha Metochi,
recent excavations of the 23rd Ephorate of Prehistoric and Classical Antiquities
and subsequently the Heraklion Ephorate of Antiquities brought to light the Cult
Centre of the city of Knossos, which is characterised by diachronic cult dating
from the Protopalatial period (18th–17th century BC) up to the 2nd century AD and
beyond. The cult in the various buildings of the plot focused on liquids, presumably water and drinking. The material from this excavation is still unpublished and
at present two volumes of the final publication are being prepared.
A series of cult buildings were found one below the other (Κάντα 2018, 252–
263). Working backwards chronologically, a double Roman Temple was discovered in the latest cultic level, surrounded by a temenos wall, with viewing steps,
an altar, a Thalassa and cult objects. It was dedicated to Artemis and Zeus, while
Serapis, Osiris and Isis were also venerated (Κάντα 2018, 252–255; Kanta 2020).
Hellenistic buildings were found underneath the Roman Temple, while a
shrine, which we named “Fetish Shrine” analogous to the one which Sir Arthur
Evans excavated at the Little Palace, was lying deeper still. It dates from the Final
Palatial Period into LM III B (ca. 1400/1380 to ca. 1230 BC). Under this shrine there
was a large Neopalatial building dedicated to cult, of which 18 spaces have been
excavated. Repositories and foundation deposits were a characteristic of this
building (fig. 1).
Room 1 included a repository of pottery and other finds as well as a “religious
sceptre” inscribed all over in Linear A. This object was presented in the recent 15th
Mycenological Conference by A. Kanta, Th. G. Palaima, D. Nakassis and M. Perna.
Another very important room of the Neopalatial building was excavated in
the northwest end of the plot. It is Neopalatial Room 3 and its finds were of outstanding significance. It is only partly excavated because most of the room lies
outside the Anetaki plot and under the baulk of the next field. Under the floor of
the room a foundation deposit was unearthed. It was covered by a round stone
slab (fig. 2).
After its removal a rough clay vase containing carbonized wood was found
and underneath were four miniature offertory double axes made one each of four
primary “precious” metals: gold, silver, bronze and iron. There were also bars of
various metals and beads (fig. 3).
The excavated part of the room produced a number of embossed gold sheets
and fragments of a gold band, which have kept their shape suggesting that they
had been fixed over a hard surface. Since no metal or stone artifacts, whole or in
fragments, were found with them, it seems that the embossed gold sheets were
fixed over a wooden object (fig. 4).
A Hieroglyphic seal from the city of Knossos
63
Fig. 1: Ground plan of the Neopalatial Building. The location of the “Fetish Shrine” is indicated
(ground plan by Konstantina Vassilopoulou).
Fig. 2: Neopalatial Room 3. The slab covering the Foundation Deposit (photo excavator’s archive)
64
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Fig. 3: Objects in the Foundation Deposit, in situ (photo excavator’s archive)
Fig. 4: Suggested use of the golden bosses (drawing by Kon
stantina Vassilopoulou)
The shape of the bosses makes it unlikely that they covered a box and suggests
the possibility that they were fixed on a wooden xoanon comparable to the clay
figure from Hagia Triadha (Paribeni 1904, 724, fig. 24).
A Hieroglyphic seal from the city of Knossos
65
The Hieroglyphic seal (Δ3702)1 which is the subject of this paper came from
the level, the removal of which exposes the floor of the room. The embossed gold
sheets and band came from this same level covering the floor. Various vases used
in cult were also found in the same level (fig. 5).
Fig. 5: Jug and chalice from the floor level of Room 3 (photos excavator’s archive)
It is interesting for chronological reasons to note that the decoration of the gold
double axe found in the foundation deposit of this room (fig. 6) is similar to that
Fig. 6: Miniature gold double axe from the
Foundation Deposit of Room 3 (photo excava
tor’s archive)
1 The seal was examined and analysed by A. Giumlia-Mair. “The analytical data, the microscopic images, and the very crisp design of the incisions all strongly suggest that the decorated seal
bead is made of steatite, cut from the bulk of the stone, shaped to bead, then incised and heated
to high temperature.” See below the Appendix.
66
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
of the double axe on the Hieroglyphic seal and on the Arkalokhori double axes
(Δημοπούλου-Ρεθεμιωτάκη 2005, 118–119). An early Neopalatial date for the Hieroglyphic seal and objects in the foundation deposit seems appropriate at the
present state of our study of the material from the Neopalatial Room 3, the “Room
of the Golden Bosses”. As mentioned above, most of Room 3, however, has not
been excavated so far. When the excavation of the rest of the room is completed,
we will be able to evaluate the complete evidence for this deposit.
The Hieroglyphic seal (T. G. P. – M. P.)
Seal KN S (4/4) 01 is an “irregular cushion seal” with four faces2. Two of the faces
are wider, rectangular and convex (α and γ) measuring 1.5 x 1.00 cm. The other
two faces are smaller, rectangular and flat (β and δ) measuring 1.5 x 0.4 cm (fig.
7). The upper and lower faces are flat, elliptical in shape and measure 1.00 cm x
0.7 max. They are perforated by a channel of 2 mm in diameter that runs through
the seal along its major axis.
Face α is divided by a horizontal line into two perfectly equal registers. Each
register measures 0.75 x 1.00 cm.
Face β is divided by two horizontal lines into three registers that measure 0.5
x 0.4 cm and its three signs have a maximum dimension of 0.3 cm.
Face α
Face β
Face γ
Face δ
Fig. 7: The four faces of the seal KN S (4/4) 01. Scale 3 : 1
2 It is defined as “irregular” by Judith Weingarten (personal communication), one of the authors
of Minoan Cushion Seals, 2014. Cushion seals are normally rectangular stones with biconvex
faces. In addition to the two convex faces (which defines a ‘cushion seal’), our seal also has two
small side faces. We thank J. Weingarten for her suggestions and her always extremely useful
comments. We also thank A. Greco and M. Hatzimanolis for the photos of the seal.
A Hieroglyphic seal from the city of Knossos
67
Face γ is divided into two registers. Each register is divided into two boxes
that measure 0.75 x 0.5 cm.
Face δ is divided into two registers that measure 0.75 x 0.4 cm. Each register
has a sign with a maximum dimension of 0.6 cm.
The signs incised into the two major faces (α and γ) have a height of between
0.5 and 0.6 cm.
The inscription3
{!} {} {} {!}
{} {!}
α1
α2
β
γ
δ
042-019
019-095-052
Δ|ϡ|ϙ|
{•}042|010|{•}
*181|*164|*165
In face γ, there are four signs that, following CHIC, are of the type “décoration
éventuellement signifiante non évidente”4.
Since the purpose of CHIC’s “transcription rules” is to avoid the creation of
many different transcription systems, depending on how each scholar considers
the signs, in this article we will follow the editors of CHIC5. Therefore, we will not
transcribe all the signs of face γ because, probably, they are not to be considered
part of the inscription stricto sensu. There has been considerable, and still continuing, debate among scholars and among us, about which signs are to be read
as signs of script proper (phonograms, logograms or metrograms). For example
M. Jasink (Jasink 2009, 137)6 observes that many of the signs that CHIC excludes
3 For the transcription of the signs we have followed the conventions of Olivier and Godart 1996,
hereafter CHIC, see p. 13–15.
4 CHIC, p. 14: “La décoration éventuellement signifiante non évidente a été traitée par nous
de deux façons différentes selon qu’elle était constituée de syllabogramme ou de logogrammes
potentiels (nous les avons alors rendu dans la transcription en caractères standardisés par le
syllabogramme ou le logogramme entre accolades, ainsi {} ou {}) ou selon qu’elle était constituée d’une représentation n’entrant pas [pour le moment au moins] dans la liste de nos signes
(nous l’avons dans ce cas rendu anonymement par {!}: on se reportera à la photo et au dessin
pour en avoir une idée); dans les deux cas, cette indication a été ignorée dans la transcription
numérique du document …”
5 Ibidem.
6 It is important to admit that there is some debate even among the authors of this article as
to whether the convention of CHIC is sound in any and all cases. Palaima does not exclude the
68
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
from signs of script proper and, therefore, not to be read (i.e. transcribed) in its
transcriptions could nonetheless have either phonetic or ideographic values7.
As in many other seals, it is possible that some signs carry a message that
presupposes a communication code other than a strictly phonetic one. In other
words, we could have signs that symbolically refer to the identity, profession or
status of the owner.
Face α
a
b
c
Fig. 8: Face α of KN S (4/4) 01; seal (a), cast (b) and drawing (c). Scale 3 : 1. (a) Photo A. Greco.
(b) Photo M. Perna. (c) Drawing M. Perna
Face α of the seal is partially eroded and did not allow for the creation of a good
cast.
Following the conventions for transcription of CHIC, the word 042-019
(A-SA) in the upper register α1 is read on the cast, from left to right, and it continues in the lower register α2 with 019-095-052 (SA-RA-NE). The three signs
in α2 are upside down (i.e. rotated 180 degrees) with respect to the upper register
possibility that the signs of face γ bear specific phonetic or logographic meaning and probably
the four signs on face γ of our seal will reopen the question.
7 According to Palaima, the most conspicuous example is the face or mask of a cat that eventually develops from Hieroglyphic through Linear A into the Linear B sign *80 /ma/. That it is a cat’s
face with eyes, cheeks and ears in Linear A is definitively proved by looking at the conveniently
collected variants in Raison and Pope 1977, 51, sign 95. On the same subject see recently Ferrara
and Weingarten 2022, 118–120. That it is a sign in Hieroglyphic is argued by Krzyszkowska 2015,
especially p. 105.
A Hieroglyphic seal from the city of Knossos
69
α1. Therefore once the viewer rotates the cast, the three signs in α2 must be read
from right to left.
The word 042-019-019-095-052 (A-SA-SA-RA-NE)8 represents a wellknown sequence, the so-called Arkhanes “formula”. In fact, this sequence of
signs is a unit of meaning (whether a word or a phrase) that appears in Linear
A on religious texts like those on libation tables, in the sequence A/JA-SA-SA-RAME9 but also on the figurine from Poros Herakliou, POR Zg 1 (Olivier, Dimopoulou, Réthémniotakis 1993) and on the silver pin from Platanos, PL Zf 1 (GORILA
4, p. 161–162).
It also appears in nine seals (and seal impressions) in Hieroglyphic (#202
ARKH S; #251 ARKH S; #252 ARKH S; #315 ARKH S; #205 CR S; #292 GOUVES S;
#179 K I; #203 KN S; #313 MONI OD. S)10. One of these documents, #205 CR S, is
also a cushion seal, but it only has a single inscribed face that is formatted in two
registers like face α of our seal KN S (4/4) 01. However, as seen in fig. 9, the signs
in the two registers are arranged with different orientation. The inscription is to
be read starting at the right of the lower register and running right to left and
continuing right to left in the upper register. As can be seen on both seals, the seal
carver does not have a boustrophedon/scripta continua instinct.
Fig. 9: Cast of seal #205 [1] CR S (1/2) 01. Not in scale (courtesy of
CMS)
It has been proposed that the presence of the same sign sequence (videlicet ‘same
word’) in both Minoan scripts (Hieroglyphic and Linear A), according to L. Godart
(Godart 1999, 299–302), should demonstrate not only that the most ancient docu-
8 Using conventionally the basic phonetic values of the corresponding signs in the Linear B
script.
9 See GORILA 5, index, p. 161–162, 228–229. See also JA-SA-SA-RA-MA-NA in KN Za 10a–b.
10 We should note here that CHIC does not assign any of the forms of sign 095 a Linear A equivalent (CHIC, 17 and 378). We do assign tentatively the conventional value RA as dictated by the
Arkhanes and Linear A “libation-table” formula.
70
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
ments in Crete, the Arkhanes seals, but that also all other seals and seal impressions that have the Arkhanes “formula” are written in Linear A. This hypothesis
consequently also involves our seal and therefore it is important to clarify whether
this hypothesis is supported by evidence now at our disposal.
First, this hypothesis is based on a single fact that the same sequence is
attested both in Hieroglyphic and in Linear A. But such an attestation is neither
significant nor surprising. The two scripts belong to the same cultural context
and even in the case of two different languages, as we are certain is the case
with the dominant languages put down in writing in Linear A and B respectively,
words common to two different scripts normally exist. Let us take for example the
word SU-KI-RI-TA attested on the Linear A nodule PH Wa 32 and on the Linear B
tablets from Knossos.
In reality, we are not dealing here with the same sign-unit because all the
attestations of the so-called Arkhanes “formula” in Linear A always have the sign
013 (ME) at the end of the sequence. On the other hand, all the attestations on
seals, which have always been considered to be in Hieroglyphic script, have at the
end of the sequence the sign 052 (NE). This is a difference of no small importance that represents a distinction between the two groups of documents. We ask
other scholars to consider whether it is significant that the two signs that end the
“formula” are both nasals, one bilabial (Linear A me) and the other dental (ne).
Second, the seals in Linear A are very rare. For only one seal we are sure that
it is written and to be read in Linear A (CR(?) Zg 4). The other three possible Linear
A seals are incised very roughly and are doubtful11. Nevertheless, let us assume,
for the sake of argument, the maximal case, namely, that these four documents
are genuinely Linear A. In this case, Linear A on seals would only constitute 0.2 %
of all Linear A documents. In contrast, the Hieroglyphic documentation on the
seals and seal impressions represent 50 % of the whole corpus.
Finally, another important difference is that in all its attestations on seals
this interesting sequence or “formula” always has sign 095 in penultimate position. Although 095 probably corresponds to the sign 060 (RA) in Linear A12, it still
maintains its own distinctive and complex pictorial shape. This indicates that the
so-called Hieroglyphic has its own peculiar and identifying repertory and this is a
significant distinction between the two scripts. In fact, in Linear A there is no sign
11 ARM Zg 1, KN Zg 55, CR(?) Zg 3.
12 In a recent contribution Perna 2019, 52–54, tried to demonstrate, through a paleographic
analysis, the derivation of the 060 sign of Linear A from the 095 sign, but this could have occurred in a very ancient phase, at the time of the creation of Linear A, probably taking a cue from
a less calligraphic variant of the 095 sign, a variant which, however, has not reached us. All the
attestations of this sign, in fact, come from seals and seal impressions.
A Hieroglyphic seal from the city of Knossos
71
in any document that vaguely looks like Hieroglyphic sign 095, not even on the
libation tables on which the Arkhanes “formula” appears. To explain this peculiar fact, Godart (Godart 1999, 300; Godart 2020, 237) suggests that Hieroglyphic
sign 095 (as well as the other signs of the Arkhanes sequence) is the archetype of
the Linear A sign 060 (RA).
Looking at forms of the sign (060) RA on the tablets of Phaistos13, which represent the oldest and most archaic group of documents in Linear A, it is evident
that in the protopalatial era the forms of the sign (060) RA had very few differences from the forms used in the Neopalatial period. Moreover, they are very
different from the sign 095 as it occurs on our seal which, according to Godart,
would be from the protopalatial period and therefore coeval with the Linear A of
Phaistos (Godart 2020, 237), an opinion that we also share.
In search of concrete elements to support this hypothesis, Godart proposes a
different interpretation for some Hieroglyphic seals (Godart 1999, 301). Regarding
seal #292 from Gouves with the Arkhanes “formula”, he believes that the four
Hieroglyphic fractions 309 (), 307 (), 308 () and 302 () appearing on the β and
δ faces of the seal can be identified respectively with the Linear A fractions A703
(), A702 (), A 713 () and A 704 () which supposedly have exactly identical
shapes. In reality this is only true for three of the four fractions.
Fraction 308 of the Hieroglyphic script does not resemble at all the fraction
A713. It is evident that Godart has confused the Hieroglyphic fraction 308 with
fraction 304 (). In fact it is this fraction that corresponds to the fraction A713 ()
in Linear A. It is quite evident, however, that there is a typo. The only possible
parallel would be fraction A 712 () which has been inserted rotated and slightly
modified (see fig. 10) in the table of comparison between the signs of the Hieroglyphic, Linear A and Linear B in CHIC, p. 19. In reality the sign as it appears
in this table, in the column of Linear A fractions, does not exist. The only two
attestations of fraction A712, as can be seen in fig. 10, are very different. But we
can also agree that a Hieroglyphic fraction consisting of a square with a linear
appendage () could correspond in Linear A to a triangle with a linear appendage
(). But if Gouves’ seal #292 has a fraction sign used only in Hieroglyphic rather
than in Linear A, this element per se demonstrates that the sequence 042-019019-095-092 belongs to the Hieroglyphic script and not to Linear A.
13 Only a fragment of a Linear A tablet from the South West Houses of Cnossos (Schoep 2002,
22, n. 43) is older than the Phaistos tablets because it dates back to the MM IIA, i.e. between 1850
and 1800 BC.
72
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Fig. 10: Detail of the table on p. 19 of CHIC
Fig. 11: GORILA microfiches, A 712, p. 277
Fig. 12: Face β and δ of the seal #292 from Gouves
The second seal analysed by Godart is Arkhanes #251. This seal also has the
Arkhanes “formula” and it has on face γ, two signs that the editors of CHIC identify, very doubtfully, with the two signs 038 () and 094 () of the Hieroglyphic
script (CHIC, 253). Starting from a possible similarity of these two Hieroglyphic
signs with signs AB 57 () and AB 38 () of Linear A, Godart proposes to consider that these two uncertain signs are signs of Linear A (Godart 1999, 301). But
observing sign 038, which would correspond to the sign AB 57 of Linear A, it is
significant to note that in Linear A the number of parallel strokes of this sign is at
most five. However, the sign on seal #251 has six or even seven strokes (fig. 13b).
Furthermore, as shown in fig. 13a–b, this attestation of sign 038 on the Arkhanes
seal #251 finds a perfect correspondence on the Hieroglyphic seal #204 of Mallia
(fig. 13a), where sign 038 is present with six parallel strokes and obviously it is a
Hieroglyphic sign as it is followed by sign 034 of the same script. Also in this case
the parallel proposed with the signs of Linear A is not convincing and indeed the
only possible parallel is with a sign of the Hieroglyphic script. Even this example,
A Hieroglyphic seal from the city of Knossos
73
therefore, goes against a possible identification of the signs of the Arkhanes
“formula” with the signs of Linear A.
As we will see later, the Hieroglyphic fraction 308 is probably present on face
β of our seal in a form similar to the only four examples existing in Cretan Hieroglyphic. Again in this case, the fraction does not resemble any Linear A fraction.
In conclusion, therefore, there are no convincing elements that can be used to
prove that all the seals with the Arkhanes “formula” or our seal KN S (4/4) 01 are
inscribed in Linear A.
Fig. 13a–b: The Hieroglyphic sign 038 on the seals #204 (a) and #251 (b) from CHIC. Not in scale
Let us now analyse the signs of the α face of our seal.
The sign 042 () appears in an accurate shape, like the other signs. The body
of the double axe has a decoration with oblique and parallel lines. The two blades
have a decoration with two slightly curved parallel lines. The best parallel is a
seal impression on a nodulus from Samothrace (#137).
a
b
c
d
Fig. 14: (a) The sign 042 on seal impression #137 from Samothrace. From CHIC, p. 192. The
sign 042 on KN S (4/4) 01; seal (b), cast (c) and drawing (d). Not in scale. (b) Photo A. Greco. (c)
Photo M. Perna. (d) Drawing M. Perna
Only sign 042 on our seal and that on the Samothrace nodulus share this particular shape and decoration; the other attestations on seals are different.
Sign 019 () in both the lower and the upper registers appears in a common
form (as, for example, on seal #205 from Crete in fig. 9).
74
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Sign 052 () appears as a vase that presents details of the rim, the neck, the
body, the handle, the tall foot and the base. Its spout is less preserved and less
visible. The handle is prominent as in seal #202.β from Arkhanes.
a
b
c
Fig. 15: Two microscope shots of the sign 052 (a) and (b). Photos of M. Hatzimanolis. On the
right (c), the sign 052 on seal #202 (CMS II, 1, 394)
We take up sign 095 () last because this new attestation is of a great importance
for the paleography of the Hieroglyphic script.
a
b
c
Fig. 16: The sign 095 on seal KN S (4/4) 01. Microscope photo, drawing and microscope photo
of the cast. (a) Photo M. Hatzimanolis. (b) Drawing M. Perna14. (c) Photo M. Hatzimanolis
All of the nine attestations of this sign are on seals or seal impressions, and
always in the so-called Arkhanes “formula”. Until now, looking at these nine
14 The drawing was made by M. Perna starting from many pictures taken by microscope or
macro photo because the sign measures only 0.5 cm in height.
A Hieroglyphic seal from the city of Knossos
75
attestations, it was difficult to understand what sign 095 might represent in its
idealized form.
M. Perna (Perna 2019, 53–54) recently called attention to the fact that in the
seal of Gouves #292, sign 095, simplified and stylized, could be compared to the
figure of a man (fig. 17b and 18) and that on seal #202β, sign 095 looks remarkably
like a profile of a human head with details of the hair, an eye, mouth and nose
(fig. 17a)15.
a
b
Fig. 17: The sign 095 in seal #202.β (courtesy of CMS) and #292.γ of CHIC (photo M. Perna)
Fig. 18: The variants of sign 095, modified, from CHIC, p. 421. The sign 095 in seals #20216 and
#315 is redesigned and in seal #252 it is corrected and overturned.
With a good lens, it was possible to see that the sign 095 on our seal KN S (4/4)
01 indeed depicts a human face and we confirmed this result with photos under
the microscope. We can also appreciate how the vase (sign 052), the human
face (sign 095) and the double axe (sign 042) are all signs engraved as realistic images with a certain sure and magnificently delicate skill. Sign 095 shows a
human head in profile with the details of the eye, the eyelids, the front, the nose,
15 Kanta and Palaima wish to acknowledge that Perna was first to observe and make sense of
these details and was able, with a high-powered microscope in Heraklion Museum, to guide
Palaima and Kanta through his analysis of the features that are now clear in the photograph and
drawing we present here.
16 In CHIC the drawing of the 095 sign of seal #202β (see fig. 22c) was made using a badly made
cast (see fig. 22b) and it absolutely does not look like the sign on the cast of the CMS II, 1, 394 (fig.
22a) nor like the sign photographed by using the microscope (fig. 20b).
76
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
the lips, the ear and the chin. An Adam’s apple seems to be represented (but a
beard is not excluded), although on seals #202 and #292 it seems to us that a long
beard is preferable (see fig. 20b–c). There is a regular pattern on the head that
seems to represent hair styling, visible also in seal #202 (see fig. 20b).
Fig. 19: The head of Mallia’s sphinx and the sign 095 on seal KN S (4/4) 01
The profile head of sign 095 shows some points of contact with the head of Mallia’s sphinx, for example the regular pattern on the head and the beard.
In light of this new attestation we can recognize, albeit in a very stylized
form, features of a human face also in other attestations of the sign.
In fig. 20 it is possible to see the new attestation of the sign 095 (a) and the
specimens of Arkhanes (#202 (b) and #252 (c)), Gouves (# 292) (d) and Moni
Odighitria (#313) (e) that we photographed under a microscope at the Heraklion
Museum.
a
b
c
d
e
Fig. 20: From left to right, the sign 095 in (a) KN S (4/4) 01, (b) #202 and (c) #252 from Arkha
nes, (d) #292 from Gouves and (e) #313 from Moni Odighitria. Photos of (b), (c), (d) and (e) are
of M. Perna
A Hieroglyphic seal from the city of Knossos
77
But even the specimens in the British and the Ashmolean Museum (#251, #205
and #203) in the light of this new attestation show a stylized human figure (fig.
21). Although the shape of sign 095 in these seals is very stylized, the main features of the sign are the same.
Seal #315 (fig. 21d) has a very eroded surface but, nevertheless, it is possible
to see that sign 095 has a profile similar to that of the other specimens (see fig.
21a–d, where the detail of sign 095 of #205 and #315 has been reoriented to the
right, to better compare the profile of the signs).
a
b
c
d
Fig. 21: Above, the seals #251 (a), #205 (b), #203 (c), #315 (d) and below a detail of sign 095 in
each seal. Photo of (a), (b), (c) courtesy of CMS. Photo of (d) M. Perna
Regarding sign 095, Ferrara et alii consider it as a bird underlining that: “we do
have variants of CH 095 which show features suggestive of tail, a leg and feathers”
(Ferrara et alii 2021, 53). In fact, even Ferrara et alii have used a misleading image.
The image of sign 095 relating to seal #202 that appears in figure 8a of their article
is a drawing published in CMS II, 1, 394b-1 which does not correspond to what is
actually preserved on the seal. The mistake was to use a drawing (which is always
an interpretation) and not the perfect cast of the seal (CMS II, 1, 394b-2), where
we can clearly see the profile of a human head with the nose, eye, hair and beard
(see supra fig. 17a and 20b). It is clear that the proposed parallel between sign 095
and the image of a bird has to be discarded.
As Perna has shown (Perna 2019, 53), the drawing of seals #252 and #315 provided in CHIC has to be corrected because a triangular filler in both cases was
considered an integral part of sign 095 (see fig. 22 below). The filler is below sign
095 in #252 and above it in #315. But this triangular filler, also clearly present in
78
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
seals #202 and #203, below sign 095 (see fig. 22a–b–c–d), is definitely not part of
this sign, as was proposed in CHIC17.
a
b
c
d
#202. Courtesy of CMS
#202 (CHIC)
#202 (CHIC)
#203. Courtesy of CMS
e
f
g
h
#252
#252. Drawing in CHIC
#315
#315. Drawing in CHIC
Fig. 22: The triangular fillers below and above sign 095
Face β
The β face has three signs each inserted in a box (fig. 24a–b–c). Starting from the
top, they are the fractions 302 ( ), 309 () and probably 308 ()18. Two of these
17 In fig. 23 it is possible to see that our seal and probably also seal #315 have a filler between
signs 095 and 052 (in both seals, rounded but very shallow), in the same position as the triangular filler on #202, #203 and #252. Unfortunately both images are not very clear and therefore
caution is a must.
18 The sign in the lowest (third) box is damaged. Fraction 308 () has only four attestations,
#206α, #291δ, #292β and SY Hf 01. The best parallel is #292 β (see fig. 24d). This identification
is not certain, in fact the sign *157 () cannot be excluded. This sign has a single attestation in
#291β where it is probably associated on the same face with fraction 308 (). Also on seal #291
(face δ) are attested the fractions 302 ( ) and 309 ().
A Hieroglyphic seal from the city of Knossos
a
79
b
Fig. 23: (a) Sign 095 in seal KN S (4/4) 01 and at its right side, the rounded filler. (b) Sign 095
in seal #315 and to its right, the rounded filler. (a) Photo of M. Hatzimanolis. (b) Photo of M.
Perna
signs, 302 ( ) and 308 (), are present together and inserted each in a box on
the face δ of seal #292 from Gouves, one of the nine documents that present the
Arkhanes “formula”. The third sign on face β, the fraction 309 (), is present on
the β face of the same seal from Gouves but in association with fraction 307 (fig.
24d–e). The same three fractions 302, 309 and 308 are present together on seal
#206 (see Table 1, A).
It is reasonable to imagine that it is no accident that the Arkhanes “formula”
and the same three fractions are present on both seals, although, in order to be
cautious, we should leave open the possibility that the fraction signs on the seals
are not used for their numeric value but rather for a symbolic meaning (see below).
a
b
c
d
e
Fig. 24: On the left, photo (a), cast (b) and drawing (c) of face β of seal KN S (4/4) 01. Scale 3 : 1.
On the right, faces δ and β of seal #292 from Gouves (d–e). Not in scale
80
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Face γ
Face γ is divided into four boxes (fig. 25). Each box has a sign. An argument can
be made that these signs are not components of true inscriptions but may have
been used as items of “décoration éventuellement signifiante non évidente” following CHIC19. In fact we may have to do with a mixing of signs in Hieroglyphic
and Linear A scripts. In the upper register of the seal, on the left there is the sign
042 (), the double axe, ligatured with a “floral appendage”. The rendering of
sign 042, the double axe, has the same carefully executed decorative features on
face γ that it has on face α.
a
b
c
Fig. 25: Face γ of KN S (4/4) 01. Seal (a), cast (b) and drawing (c). Scale 3 : 1. (a) Photo A. Greco.
(b) Photo M. Perna. (c) Drawing M. Perna
In the box on the upper right of the seal, there is a sign, which recalls sign AB 77
() of Linear A, that (until now) was not attested in Hieroglyphic. It is ligatured
with the same “floral appendage”. Inside AB 77 there are two wavy perpendicular
strokes. AB 77 appears in such a form in Linear A on the libation tables of Kophinas (Za 1c) and Iouktas (Za 6) and on a roundel from Phaistos (Wc 37a) and even
in Linear B tablets from Pylos (Palaima 1988, 252). Another sign that looks like AB
19 CHIC, p. 13–14. As we discussed above (see notes 6 and 7), there are some serious problems
with taking signs that depict images in a Hieroglyphic or pictographic text as mere space fillers
or decorative motifs. On face γ of KN S (4/4) 01, the “floral appendage” that appears in three of
the registers is not far removed from the certain Linear A sign AB 04 () as drawn accurately in
GORILA 5, p. XXVIII (see especially PH Zb 4 and KO Za 1d). We should note, however, that the
Linear AB 04 sign has its central line always vertical. On our seal the central stroke is curved and
the sign is inclined. Below we discuss how this “appendage” exists in Hieroglyphic in ligature
with other signs.
A Hieroglyphic seal from the city of Knossos
81
77 appears on the seal impression #135 on a roundel from Samothrace (fig. 26a).
In this case, following CHIC, it seems to be used as a decoration, as it is on many
painted vases of the Mycenaean palatial period. This roundel, probably, presents
on the recto the Linear sign AB 081 ().
On a seal (CMS VI, 147), not included in CHIC, there are at least two of the
signs on the γ face of our seal (see below fig. 26b–c). The first is a circle with
an internal cross (therefore a sign that formally has the outline of sign AB 77)20
which has the exact same floral appendage as on our seal. The second sign is the
010 sign of the Hieroglyphic script which has an appendage made up of a curved
line and some “balls” that look like the appendage of our seal.
a
a
b
b
c
c
Fig. 26: On the left (a), seal #135 of CHIC. On the right seal CMS VI, 147, seal (b) and cast (c)
We should point out, too, that Hieroglyphic sign 027 recalls the “floral appendage” here (fig. 27c). A similar appendage is incised on a nodulus from Samothrace
(SA We 4, fig. 27a–b), but the sign ligatured to it is not a true double axe (the
standard shape of this sign in Hieroglyphic). The sign on SA We 4 is very stylized
and resembles sign AB 08 (A) in Linear A more than the standard Hieroglyphic
sign. A second Linear A inscription AB 06-04 () is incised on another nodulus
from Samothrace (SA We 3) impressed with the Hieroglyphic seal bearing the
inscription 042-019 (CMS V, Suppl. 1B, 327).
However, we should not make the mistake of thinking that the Hieroglyphic
inscription and the Linear A inscription on the same Samothrace documents may
have some relationship. In fact, the two inscriptions were created independently
and are distant in space and time21. The Hieroglyphic seals impressed on the
Samothrace documents may have been made in a time long before the Linear A
inscription (AB 06-04) and probably somewhere else too. Furthermore the same
20 This attestation of the possible sign AB 77 was indicated to M. Perna by B. Montecchi.
21 About Samothrace documents see Matsas 1991, 159–179 and Matsas 1995, 235–247 and pl.
XXIV–XXVIII.
82
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
seal may have sealed, over time, roundels and nodules bearing records of various
kinds. It is therefore evident that no textual relationship can safely be deduced
between the two inscriptions.
a
b
c
Fig. 27: (a)–(b) The nodulus SA We 4 (CMS V Suppl. 3,2 n° 343). Not in scale. Courtesy of CMS.
On the right (c), the Hieroglyphic sign 027 from CHIC, p. 395
In the lower registers of face γ of our seal (fig. 25) there are two signs, each in
a box. On the left there is a sign that reminds us of the sign AB 41 of Linear A
(or simply a trident) ligatured with the same “floral appendage” as in the upper
register (fig. 25). A “floral appendage” is ligatured with a sign that has been identified, but uncertainly, with the Hieroglyphic sign 049 () in seal #259.α (fig. 28).
Fig. 28: Seal #259.α from CHIC, p. 256
The sign on the right, in the lower register, is the Hieroglyphic sign 010. The best
parallel for this sign is on seal #260.β (fig. 29b). The empty space next to sign 010,
in our seal, is filled by six “dots” that form a curved line following the contour of
the outer line to their right22. As we have seen, in seal n° 147 CMS VI there is the
same appendage with some “dots” or “balls” (see above fig. 26b–c).
Sign 010 is also present in face L of seal #315 of CHIC (CMS II, 1, 391L; see
fig. 29c)23. This is one of the seals with the Arkhanes “formula” and this demon-
22 This filler recalls the logogram *171 () of the Hieroglyphic script which is attested only on
clay documents.
23 But in this case the fillers are not dots but small triangles.
A Hieroglyphic seal from the city of Knossos
83
strates that a relationship between the “leg” and the Arkhanes “formula” probably exists.
a
b
c
Fig. 29: The sign 010 in face γ of KN S (4/4) 01 (a), in #260.β (b) and in CMS II, 1, 391L. Courtesy
of CMS (c)
Face δ
Face δ (fig. 32) has two signs, each in a box. The upper box has the logogram
*181 () that has been attested only thus far on seal #305 from Lastros (fig. 30a),
in association with logogram*180 . Sign *181 is attested also on seal #315 from
Arkhanes (fig. 30b) but, following CHIC, with no writing function as well as on
seal CMS II, 1, 392a, again from Arkhanes (fig. 31b), and on seal CMS II, 1, 126
from Kalathiana (fig. 31a). Sign *181 is isolated on these three seals. If we follow
CHIC, we would conjecture that it is not used as a sign of writing. In this case we
would think that this ‘sign’ had a symbolic value of some importance, conveying
an evident message in the Minoan world but one that escapes us24. However, we
have decided to transcribe the sign as *181, because on our seal it is not isolated
but is rather followed by another ideogram as is the case of the seal from Lastros
(#305) the signs of which are transcribed in CHIC (p. 285).
The lower box of face δ (fig. 32a–b–c) has a sign never attested until now in
Cretan Hieroglyphic. The only convincing parallel for this sign is the Hieroglyphic
logogram *164/*165, corresponding to the AB 180 logogram of Linear A. It repre-
24 Even if, correctly, from Evans onwards scholars have tried to recognise a meaning in these
isolated signs, we must not forget that these signs could have been inserted, both to fill any
spaces left after writing the text proper or for purely decorative purposes. But one hypothesis
does not exclude the other, because in filling these spaces one could use signs that in one way or
another “spoke” about the function or the rank of the seal owner.
84
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
a
b
Fig. 30: (a) The signs *181 and *180 on the Lastros seal #305 from CHIC. (b) The sign *181 on the
Arkhanes seal #315 (photo M. Perna)
a
b
Fig. 31: (a) The seal from Kalathiana CMS II, 1, 126. (b) The seal from Arkhanes CMS II, 1, 392a
(courtesy of CMS)
sents a skin or hide of an animal. The Hieroglyphic sign is attested in bar #48 from
Knossos25 in the two variants *164 and *165 that also exist in Linear A (fig. 32d)26.
a
b
c
d
Fig. 32: On the left, face δ of KN S (4/4) 01. Seal (a), cast (b) and drawing (c). Scale 3 : 1. On the
right, signs *164 and *165 on the bar #048 from Knossos (d). (a) Photo A. Greco. (b) Photo M.
Perna. (c) Drawing M. Perna. (d) From CHIC, p. 98
25 For bar #48 see CHIC, p. 98–99. For the ligatured forms of AB 180, see GORILA 5, p. 280.
26 It must be remembered that the editors of CHIC, p. 18, consider that the Hieroglyphic bar #48
from Knossos could actually be written in Linear A.
A Hieroglyphic seal from the city of Knossos
85
A form without appendices is not known until now in Hieroglyphic script and
the sign on our seal has also oblique and parallel lines inside; in this case, it is
also probably an iconographic motif depicting a skin rather than a script sign per
se. But we stand to be corrected by other scholars or by future finds.
Conclusions
The Hieroglyphic seal KN S (4/4) 01 is a document of considerable importance in
the study of Aegean script for four main reasons. First, with regard to iconography and palaeography, we have identified for the first time the true idealized form
of sign 095 (fig. 16a–c) and that permits us to correct our understanding of the
shape of other examples of the same sign and thereby to improve our knowledge
of the palaeographical development of the Hieroglyphic script.
Second, it preserves one more attestation of the so-called Arkhanes “formula”.
Third, the presence of the same fractions on seal #292 from Gouves shows a
connection between the words represented by the Arkhanes “formula” and the
fractions. In addition to the seals already mentioned, the fractions are well represented also on seals #206 and #291 (table 1 A and B)27.
On the basis of this last observation, perhaps it is useful to explain why fractions are engraved on seals even if they do not appear to be used for their mathematical values.
Fractions used on seals are never associated with numbers, as they are frequently and quite logically in economic texts. They also do not appear in the
context of logograms for food, a frequent use of fractions in true accounting texts.
On seals, fractions are often framed in boxes, seemingly explicitly separated from
the text. All this justifies the suspicion that they do not appear to be used for their
mathematical values per se.
It seems to us, in fact, not far-fetched that a scribe or an official responsible for a warehouse, since he used arithmetic for his job, could be identified on
his seal through the signs normally used in his job. We must keep foremost in
our mind that in the minds of the Minoans, fractions, which for us are abstract
mathematical values, also represented real vessels with established capacities
(or weights), used in the distribution of food, an operation of high ideological
27 The two seals #266 and #301 both have the fraction 309 but it was used as a filler, inserted,
that is, to fill an empty space (table 1, D and E). But since this fraction has the shape of a spiral,
we can not even be sure that it is really a fraction and not merely a spiral pattern.
86
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
and practical impact in a society like the Minoan palatial society that was highly
stratified, both politically and economically28.
The presence of fractions and one or even two logograms for food such as on
seals #291 (with figs) and #206 (with figs and wine) could indicate, with a more
explicit iconography, that the person in question was involved with the distribution of these products (see table 1 A and B).
Since the Arkhanes “formula” is used in cult environments in the Linear
A texts, individuals identified through this sequence on their seals could work
in cultic contexts. Perhaps the man depicted on the K face of seal #315 from
Arkhanes (fig. 33) could be one of those individuals to whom the fractions on the
seal of Gouves #292 and on the new seal KN S (4/4) 01 refer iconically.
Fourth and finally, in addition to the relation between fractions and Arkhanes
“formula”, this seal confirms also a relationship between this “formula” and the
logogram *181 on the one hand and the sign 010 on the other hand.
Fig. 33: Face K of the seal #315 of CHIC (CMS II, 1, 391K). Courtesy of
CMS
Table 1: Hieroglyphic seals with fractions
A
28 We need only recall the importance of the root *kerh3 conveying the idea of satisfying the
general population through providing them with food to consume in the invented official palatial nomenclature of the Mycenaean palatial centres: da-mo-ko-ro and (po-ro-)ko-re-te.
Palaima 2012, 348–350.
A Hieroglyphic seal from the city of Knossos
87
B
C
D
E
Bibliography
CMS = Corpus der minoischen und mykenischen Siegel, I. Pini (ed.), Berlin.
CHIC = Corpus Hieroglyphicarum Inscriptionum Cretae, par JeanPierre Olivier et Louis Godart
avec la collaboration de JeanClaude Poursat, Paris – Rome 1996.
ΔημοπούλουΡεθεμιωτάκη, Ν. 2005. Το Αρχαιολογικό Μουσείο Ηρακλείου, Εκδόσεις Ιδρύματος
Ι. Σ. Λάτση, Αθήνα.
Evans, A. 1909. Scripta Minoa. Vol. I. The Hieroglyphic and Primitive Linear Classes [SM I],
Oxford.
88
Athanasia Kanta, Thomas G. Palaima, Massimo Perna
Ferrara, S., Montecchi, B., and Valério, M. 2021. What is the ‘Archanes formula’? Deconstructing
and Reconstructing the Earliest Attestation of Writing in the Aegean, The Annual of the
British School at Athens, 116, 43–62.
Ferrara, S. and Weingarten, J. 2022. Cretan Hieroglyphic Seals and Script: A View from the East,
Pasiphae 16, 111–121.
Godart, L. 1999. L’écriture d’Arkhanès: hiéroglyphique ou linéaire A ?, in MELETEMATA. Studies
in Aegean Archaeology presented to Malcolm H. Wiener as he Enters his 65th Year,
299–302.
Jasink, A. M., Dionisio, G. and Weingarten, J. 2014. Minoan Cushion Seals. Innovation in Form,
Style and Use in Bronze Age Glyptic, Rome.
Κάντα, Α. 2018. Το Θρησκευτικό Κέντρο της Πόλης της Κνωσού και η διαχρονική του λατρεία.
Κρήτη. Αναδυόμενες πόλεις | Κατάλογος έκθεσης Άπτερα ― Ελεύθερνα ― Κνωσός. Τρεις
αρχαίες πόλεις ζωντανεύουν, Ν. Χρ. Σταμπολίδης, Ε. Παπαδοπούλου, Ι. Γ. Λουρεντζάτου, Ι.
Φάππας (eds.), Αθήνα, 250–265.
Kanta, A. 2020. A “thalassa” at Knossos”, in Eleutherna, Crete and the Outside World.
Proceedings of the International Archaeological Conference, 31 May – 2 June 2018,
Grecotel Creta Palace, Rethymno and 3 June 2018, Museum of Ancient Eleutherna, N. Chr.
Stampolidis and M. Giannopoulou (eds.), Athens – Rethymno, 405–410.
Kanta, A., Nakassis, D., Palaima Th., Perna, M. and Weingarten, J. forthcoming. Two Linear A
Inscriptions on Ivory (KN Zg 57 and KN Zg 58) from the Cult Centre of the City of Knossos.
Karnava, A. 2016. On Sacred Vocabulary and Religious Dedications. The Minoan “Libation
Formula”, in E. AlramStern et alii (eds.), Metaphysis. Ritual, Myth and Symbolism in the
Aegean Bronze Age. Proceedings of the 15th International Aegean Conference, Vienna,
22–25 April 2014 (Aegaeum 39), Liège, 345–355.
Krzyszkowska, O. 2015. Why were Cats different? Script and Imagery in Middle Minoan II
Glyptic, in C. F. Macdonald, E. Hatzaki and S. Andreou (eds.), The Great Islands. Studies of
Crete and Cyprus presented to Gerald Cadogan, Athens, 100–106.
Matsas, D. 1991. Samothrace and the Northeastern Aegean: The Minoan Connection, Studia
Troica 1, 159–179.
Matsas, D. 1995. Minoan LongDistance Trade: a View from the Northern Aegean, in R.
Laffineur and W.D. Niemeier (eds.), Politeia. Society and State in the Aegean Bronze Age.
Proceedings of the 5th International Aegean Conference, University of Heidelberg, Archäologisches Institut, 10–13 April 1994 (Aegaeum 12), 235–247.
Olivier, J.P., Dimopoulou, N. and Réthémiotakis, G. 1993. Une statuette en argile MR IIIA de
Poros/Irakliou avec inscription en linéaire A, BCH 117, 501–521.
Olivier, J.P. (forthcoming). Documents in Mycenaean Greek, 3rd edition.
Palaima, T. G. 1988. The Scribes of Pylos, Incunabula Graeca LXXVII, Rome.
Palaima, T. G. 2012. Security and Insecurity as Tools of Power in Mycenaean Palatial Kingdoms,
in P. Carlier, C. de Lamberterie, M. Egetmeyer, N. Guilleux, F. Rougemont and J. Zurbach
(eds.), Études Mycéniennes 2010, Pisa – Rome, 345–356.
Paribeni, R. 1904. Ricerche nel sepolcreto di Hagia Triada presso Phaestos, Monumenti Antichi
14, 677–756.
Perna, M. 2019. A Seal in the British Museum with a Cretan Hieroglyphic Inscription (CR (?) S
(1/1) 07), Kadmos 58, 49–59.
Raison, J. and Pope, M. 1977. Index transnuméré du linéaire A, BCILL 11.
Schoep I, 2002. The Administration of Neopalatial Crete. A Critical Assessment of the Linear A
Tablets and their Role in the Administrative Process, Suplementos a Minos 17, Salamanca.
Appendix: The seal Δ3702
89
Appendix
The seal Δ3702: examination and analysis results
(Alessandra Giumlia-Mair)
1 Introduction
The small seal bead Δ3702 was first thought to be made of bone, mainly because
of its beige or cream colouring. After cleaning the object was examined under the
microscope and it presents a very compact and relatively uniform surface with
some dark spots and reddish coloured areas (fig. 1). However no regular structure
that might have indicated an organic origin could be detected anywhere on the
artefact (fig. 2).
The rectangular plaque was examined several times under the microscope
and more microscope images were provided by Prof. Athanasia Kanta, then the
object was analysed by XRF to determine its composition.
Fig. 1: Seal bead Δ3702 from Knossos
(photo A. GiumliaMair, scale 3 : 1)
Fig. 2: Detail of surface. A slight vitrification of
the surface can be recognized. The decora
tions are deeply carved. The irregular blackish
spots are natural impurities of the stone
(photo M. Hatzimanolis, not in scale)
90
Alessandra GiumliaMair
2 Methods of examination
For the first examination several optical devices with magnification 5 X, 10 X, 12
X were employed. Then a digital microscope with interchangeable lenses (10 X
and 50 X magnification) was used before the analysis by XRF. Finally, an optical
microscope was employed as well.
For the actual analysis a transportable X-ray fluorescence spectrometry
equipment was used. This method is well known in archaeology (cf. Lutz and
Pernicka 1996; Helmig et al. 2007; Longoni et al. 1998; Mendoza Cuevas and
Perez Gravie 2011; Giumlia-Mair et al. 2018) and has been used and improved for
decades (Hahn-Weinheimer et al. 1995; Lutz and Pernicka 1996).
The main advantage of this method is that XRF is a non-destructive method,
and that a large number of elements can be determined at once. Although XRF
is a surface analysis, various experiments carried out in the past by comparing
the results of XRF and AAS analyses (AAS is considered to be much more precise,
because for this kind of analysis destructive samples taken with a drill from the
internal part of the objects are necessary, and because in this way possible alteration of the surface are avoided) have demonstrated that over 90 % of the XRF
results were well within + 20 % of the corresponding AAS results, when the work
was carried out by experienced analysts and proper standards were used. The
equipment employed for this study had been specially developed for the analysis
of archaeological and Cultural Heritage artefacts. It can also be employed for the
analysis of very large, bulky or heavy items that cannot be easily brought to a laboratory, in museums or on excavation, and it consists of an X-ray source, a transformer, a stabilizer and a tripod with several devices that control stability and
position, as well as a computer with dedicated software. The spectrometer has
a Si(Li) detector and operates at a maximum current of 0.35 mA and a maximum
voltage of 50 kV. The measurements are performed by pointing the X-ray on a
small, flat, cleaned area on the object for a short time (typically 3–5 minutes). If
necessary, however, the measurement time can be longer, for instance, when the
spot analyzed is very small. The X-rays are emitted by a miniaturized X-ray tube.
The irradiated area has a diameter of around 1.5–2 mm, but the built-in collimator
can reduce or enlarge the area of analysis, as required by the size of the object,
the detail to be analysed (for instance an inlay or a small area free of corrosion or
other kind of alteration) and the surface texture. The measurements of the objects
must be taken at a fixed angle and from a fixed distance, and built-in devices
control these features. When the position is correct, an audio signal rings, and a
laser pointer indicates the exact spot of measurement. The quoted detection limits
for some of the most significant elements are (given in ppm): Cu 0.1; Pb 200; Sn
200; Fe 250; Co 200; Ni 150; As 5; Sb 510; Ag 200. It is important to note that light
Appendix: The seal Δ3702
91
elements and in particular Si cannot be determined with this method, so that in
this case some important elements of faience, glass, steatite etc. cannot be identified. Furthermore, the equipment employed was calibrated for metal objects, and
no standards for vitreous materials were available, therefore the analysis results
can only be considered qualitative and indicative. For more precise and detailed
analyses regarding trace elements or the structure of the material other types of
methods should be used.
Nevertheless, the optical examination, combined with the XRF analysis data
gives precious information on the material.
3 Discussion of results
The appearance of the surface and the aspect of the decoration strongly suggested
at first sight that the material out of which the object is made might be faience,
glass paste or even glazed steatite.
Vitreous materials were known and employed in the Aegean already around
the mid third millennium BC (Panagiotaki 2008, 35), especially in Crete, where
many examples have been found. Rectangular beads made of faience come for
example from the MM II Quartier Mu at Malia (Detourney et al. 1980, 133–134, fig.
186).
The XRF analyses evidenced high Ca, very noticeable but not high iron, traces
of manganese, cobalt, magnesium, and alumina.
We must keep in mind that this analysis method cannot determine silica,
which is the main component of vitreous materials, i.e., glass, glass paste and
glazed materials such as faience and steatite. Therefore, the identification of the
material, if based only on the analysis results, is rather difficult. Nevertheless, if
we take into account the results of the various optical examinations, we can reach
some sound conclusions.
Faience production in ancient times
As we know, in ancient times there were three different possibilities for producing faience beads (Giumlia-Mair and Soles 2013, 116). The first step was the production of the core, either done by hand or by using an open mould, however
the following production steps differed depending on the production areas. The
method used in Egypt was the efflorescence technique, for which quartz, soda
and a colouring agent were ground and then mixed with water. This mixture was
left to dry for a while and during the drying process the water-soluble alkali salts
present in it began to migrate to the surface, creating thus a whitish layer. When
the layer reached a sufficient thickness, the beads were fired at a temperature
92
Alessandra GiumliaMair
of up to 950° C. Subsequently the individual pieces were polished, carved and
drilled. The beads obtained in this way are characterised by a thick glaze under
which a similarly thick, but irregular layer of glass interface can be recognised.
When the direct application method was employed, the core was made of clay
and it was then covered with a layer of a paste consisting of quartz, calcium carbonate and a colouring agent. The firing temperature in this case went from 850°
to 980° C. Better results could also be achieved by employing a fritte, instead
of the raw mixture. The fritte was prepared by mixing the same ingredients and
heating them to around 800° C to obtain a glassy lump that could be reduced to
a finely ground powder. This powder was then mixed with water and an organic
glue, such as tragacanth gum, and applied on the surface of the beads that had
to be again heated to around 850° C. The result was a much thinner, but regular
and shiny glaze, almost without interstitial glass underneath. Beads produced by
this technique often show on the surface a kind of flow lines and sometimes even
some drips. Faience beads with a clay core were used in the Near East, Cyprus
and the Levant from the mid-2nd millennium onwards (Floreano and Giumlia-Mair
2008; Caubet and Pierrot-Bonnefois 2005, 66; Matoïan and Bouquillon 1999).
The third method was the cementation technique. In this case the core of the
beads was made of clay, on which a paste consisting of quartz powder, plant ash,
charcoal and a colouring agent was applied. The firing temperature was around
980° C. The glaze obtained in this way is rather thin and irregular with only a
thin layer of interstitial glass underneath, but no interstitial glass at all inside
the core. This kind of technique was instead employed in later times (Tite et al.
1983; Rehren 2000, 145; Shortland and Tite 2000; Rehren 2001; Tite and Shortland
2003).
The colour of the beads is a light creamy beige, and the microscopic examination did not reveal any layer of coloured glaze or remains of a glaze that would
be expected if the material was faience or glass paste. While the presence of iron,
manganese and cobalt determined by the XRF analyses could be taken as indication that these were the colouring agents of the glaze, their diffusion on the bead
seems to be very irregular and more like that of impurities in a natural stone.
Furthermore, the surface structure seems to show the signs of heating, but the
decorations have clearly been carved, as they are rather well defined and sharp
(cfr. fig. 2). All these details suggest that the material employed for the seal was
steatite.
Steatite working and finishing
Steatite is a natural stone, whitish to grey, sometimes green or reddish, with a
greasy or soapy feel. The name steatite comes from the Greek word στέαρ, στέατος
“fat”. This material is also often called soapstone or potstone. The family is that
Appendix: The seal Δ3702
93
of hydrous magnesium silicates and includes talc, serpentine and meerschaum.
Steatite is a variety of talc, a metamorphic type of stone, consisting of magnesium
rich talc with varying amounts of chlorite, albite and amphiboles with traces of
Fe oxides, CaO and Al2O3, found as transformation product of serpentine in crystalline schist. This lithotype is characterised by the grade 1 of the Mohs hardness
scale and can be easily cut with a knife and scratched with a fingernail (Read
1962, 408–412). Because of this characteristic it can be very easily carved and has
been widely used for the production of beads by many cultures. For instance, steatite beads were employed since the 5th millennium BC at Mohenjo Daro (Vidale
1989, 291), but also at Harappa in Pakistan (Kenoyer 1997), in Egypt (Beck 1934),
in Oman (Rosch et al. 1997), in Galilee and in various Arabic countries (Frifelt
1991; Bar-Yosef Mayer et al. 2004; Panei et al. 2005). The manufacturing process
involved several stages of cutting, shaping, chipping, grinding, perforation, polishing and carving until the desired shape and size were obtained. The blanks
were perforated using the well-known method of stone drills with a thin point
and abrasive powders (Kenoyer and Vidale 1992) or, for example, with a thin
metal point. The strand of beads could sometimes be shaped by rolling or grinding on a suitable stone slab and the edges were often faceted. When the desired
result was achieved and completed with incisions the beads were fired at around
1000° C for several hours, until the talc present in the bulk became hard and
compact (Vidale 1987; 1989; 1995). In some cases, a powdered fritte was applied
on the surface and fired as was done in the faience production to produce a glaze
or the incisions could be filled with some colouring agent to obtain a pleasing
contrast (Tite and Bimson 1989). A good example of glazed steatite is for example
the Egyptian Bes and udjat-eye plaque in the Brooklyn Museum (fig. 3), dated
to 664–332 BC, belonging to the Charles Edwin Wilbour Fund, n. 37.1337E. The
colour of this small item (it measures only 0.5 x 0.9 x 1.4 cm) is dark blue and was
certainly intended as an imitation of lapislazuli.
Fig. 3: Small glazed plaque made of steatite, with an udjateye on one side and
the figure of the Egyptian god Bes on the other. Brooklyn Museum, dated to
664–332 BC, Charles Edwin Wilbour Fund, n. 37.1337E. The colouring agent of
the glaze was cobalt (photo: Brooklyn Museum, CUR.37.1337E_view1.jpg, not in
scale)
The deep blue nuance of the glaze was in this case conferred by cobalt, which is a
very powerful colouring agent. Already 0.5 % of cobalt gives to glass and other vitreous materials a dark blue colour and it was later, for example in Roman times,
even used in the production of black glass. The glazing of steatite was widely
94
Alessandra GiumliaMair
employed in Egypt, for example in the production of the light blue glaze found on
the scarabs. For such a lighter blue colouring copper salts were employed instead
of cobalt. On the seal bead Δ3702 from Knossos, however, no traces of glaze could
be identified.
The reason why steatite becomes hard when fired at around 900° C (Vidale
1987) or more (Bouquillon et al. 1995) is that at these temperatures it decomposes
and recrystallizes in the form of artificial enstatite and cristobalite (MgSiO3) with
a whitish colour. Natural enstatite is very hard (Mohs hardness scale 5–6; Deer et
al. 1992) and cannot easily be shaped or incised, therefore it cannot be the original material employed for cutting and carving our seal bead Δ3702.
4 Conclusions
The microscopic examination of the bead showed a slight vitrification of the
surface (cfr. fig. 2), but no remains of glaze, even less of a coloured glaze, could
be found. The piercing of the seal was carefully examined as well, to check if any
differences between the outer surface and the inside of the bead could be recognized, however, none could be noticed (fig. 4).
Fig. 4: Detail showing the neatly drilled piercing on the seal bead. No difference can be seen
between the surface and the internal material (photo A. GiumliaMair, not in scale).
Seen through a microscope the surface of the bead shows tiny dark specks that
are the impurities containing iron, manganese, and cobalt, the elements identified by the XRF analysis of the bead (cfr. fig. 2). These irregular inclusions testify
that this material is a natural stone and are rather often seen, for instance, on
unglazed Egyptian scarabs made of steatite hardened by heating. Manganese and
cobalt are chemically closely related to iron and common impurities in iron, but
especially and in mafic and ultramafic igneous rocks that during metamorphism
become serpentine, steatite or chlorite schist (Vhay 1979, 4). In this case the small
Appendix: The seal Δ3702
95
amounts of the elements Mn and Co identified by XRF should not be interpreted
as colouring agents of a glaze, but as natural impurities of steatite.
The analytical data, the microscopic images, and the very crisp design of the
incisions all strongly suggest that the decorated seal bead is made of steatite,
cut from the bulk of the stone, shaped to bead, then incised and heated to high
temperature.
Bibliography
BarYosef Mayer D., Porat N., Gal Z., Shalem D., Smithline H., 2004, Steatite beads at Peqi’in:
long distance trade and pyrotechnology during the Chalcolithic of the Levant, Journal of
Archaeological Science, 31, 493–502.
Beck H. C., 1934, Notes on glazed stones. Part I – Glazed steatite, Ancient Egypt and the East
1934, 69–88.
Bouquillon A., Barthelemy de Saizieu B., Duval A., 1995, Glazed steatite beads from Merhgarh
and Nausharo, Material Issues in Art and Archaeology 4/352, 527–538.
Caubet A., PierrotBonnefois G., 2005, Faïences de l’antiquité de l’Égypt à l’Iran, Paris.
Deer W. A., Howie R. A., Zussman J., 1992, An introduction to the rock-forming minerals,
Longman, Harlow UK.
Detournay B., Poursat J.C., Vandenabeele F., 1980, Fouilles executées à Mallia: le quartier Mu,
Vol. II, Études Crétoises XXVI, École Française d’Athènes, Athènes.
Floreano E., GiumliaMair A., 2008, A small LBA faience jug from Cyprus in the Civici Musei
di Storia ed Arte, Trieste, Italy, Report of the Department of Antiquities Cyprus, Nicosia,
209–226.
Frifelt K., 1991, The Island of Umm AnNar, Jutland Logical Society Publications XXVI, 1, 112–123.
GiumliaMair A., Betancourt P. P., S. Ferrence S. C., 2018, Arsenic in the network: arsenical
copper in Minoan Crete, in GiumliaMair A. and Lo Schiavo F. (eds.), Bronze Age Metallurgy
on Mediterranean Islands. In Honour of Robert Maddin and Vassos Karagheorgis,
DrémilLafage, 520–541.
HahnWeinheimer P., Hirner A., WeberDiefenbach K., 1995, Röntgenfluoreszenzanalytische
Methoden – Grundlagen und praktische Anwendung in den Geo-, Material- und Umweltwissenschaften, Braunschweig, Wiesbaden.
Helmig D., Jackwerth E., Hauptmann A., 2007, Archaeometallurgical fieldwork and the use of a
portable Xray spectrometer, Archaeometry 31 (2), 181–191.
Kenoyer J. M., Vidale M., 1992, A new look at stone drills of the Indus Valley tradition, Material
Issues in Art and Archaeology III, Vandiver P., Drozik J. R., Wheeler G. S. and Freestone I.
(eds.), Pittsburgh, 459–518.
Kenoyer J. M., 1997, Trade and technology of the Indus Valley: new insights from Harappa,
Pakistan, World Archaeology 29 (2), 262–280.
Longoni A., Fiorini C., Leutenegger P., Sciuti S., Fronterotta G., Strüder L., Lechner P., 1998,
A portable XRF spectrometer for nondestructive analyses in archaeometry, Nuclear
Instruments and Methods in Physics Research A 409, 407–409.
Lutz J., Pernicka E., 1996, EDXRF analysis of ancient copper alloys, Archaeometry 38 (2),
313–323.
96
Alessandra GiumliaMair
Matoïan V., Bouquillon A., 1999, La céramique argileuse à glaçure de Ras ShamraOugarit, Syria
76, 57–81.
Mendoza Cuevas A., Perez Gravie H., 2011, Portable energy dispersive Xray fluorescence and
Xray diffraction and radiography system for archaeometry, Nuclear Instruments and
Methods in Physics Research A 633, 72–78.
Panagiotaki M., 2008, The technological development of Aegean vitreous materials in the
Bronze Age, in Jackson C. M. and Wager E. C. (eds.), Vitreous Materials in the Late Bronze
Age Aegean, Sheffield, Studies in Aegean Archaeology, 34–63.
Panei L., Rinaldi G., Tosi M., 2005, Investigations on ancient beads from the Sultanate of Oman
(Ra’s alHadd – Southern Oman), Archaeosciences 29, 151–155.
Rehren T., 2000, Rationales in old world base glass compositions, Journal of Archaeological
Science 27, 1225–1234.
Rehren T., 2001, Aspects of the production of cobaltblue glass in Egypt, Archaeometry 43 (4),
483–489.
Rosch C., Hock R., Schussler U., Yule P., Hannibal A., 1997, Electron microprobe analysis and
Xray diffraction methods in archaeometry: investigations on ancient beads from the
Sultanate of Oman, European Journal of Mineralogy 9, 763–783.
Soles J., GiumliaMair A., 2018, Metallurgical habits and workshop remains in LM IB
Mochlos, EastCrete, in GiumliaMair A. and Lo Schiavo F. (eds.), Bronze Age Metallurgy
on Mediterranean Islands. In Honour of Robert Maddin and Vassos Karagheorgis,
DrémilLafage, 498–519.
Tite M. S., Freestone I. C., Bimson M., 1983, Egyptian faience: an investigation of the methods
of production, Archaeometry 25, 17–27.
Tite M. S., Bimson M., 1989, Glazed steatite: an investigation of the methods of glazing used in
Ancient Egypt, World Archaeology 21, 87–100.
Tite M. S., Shortland A. J., 2003, Production technology for copper and cobaltblue vitreous
materials from the New Kingdom site of Amarna – a reappraisal, Archaeometry 45 (2),
285–312.
Vhay J. S., 1979, Cobalt in the United States, US Geological Survey, OpenFile 791436, Dept of
the Interior, Washington, 1–26.
Vidale M., 1987, Some observations and conjectures on a group of steatitedebitage concen
trations on the surface of Moenjodaro, Annali dell’lstituto Universitario di Napoli 47/2,
113–129.
Vidale M., 1989, Early Harappan steatite, faience and paste beads in a necklace from Mehrgarh
Nausharo (Pakistan), East and West 39, 1–4, 291–300.
Vidale M., 1995, Early beadmakers of the Indus tradition. The manufacturing sequence of talc
beads at Mehrgarh in the 5th millennium B.C., East and West 45, 1–4, 45–80.