The Transformation of Greek Amulets in Roman Imperial Times 0812249356, 9780812249354

Citation preview

Empire and After Series Editor: Clifford Ando A complete list of books in the series is available from the publisher.

The Transformationof GreekAmulets in Roman Imperial Times

ChristopherA. Faraone

PENN

Universityof PennsylvaniaPress Philadelphia

To myfriends Roy Kotansky,Attilio Mastrocinque,and Arpad Nagy, whosharemyfascinationwith amulets

Copyright© 2018 University of Pennsylvania Press All rights reserved. Except for brief quotations used for purposes of review or scholarly citation, none of this book may be reproduced in any form by any means without written permission from the publisher. Published by University of Pennsylvania Press Philadelphia, Pennsylvania 19104-4112 www.upenn.edu/pennpress Printed in the United States of America on acid-free paper 10987654321 Library of Congress Cataloging-in-Publication Data Names: Faraone, Christopher A., author. Title: The transformation of Greek amulets in Roman imperial times I Christopher A. Faraone. Other titles: Empire and after Description: 1st edition.

I Philadelphia

: University of Pennsylvania Press, [2018] I Series:

Empire and After I Includes bibliographical references and index. Identifiers: LCCN 2017046034 I ISBN 978-0-8122-4935-4 (hardcover : alk. paper) Subjects: LCSH: Amulets, Greek. I Magic paraphernalia-Greece. Classification: LCC DF129 .F37 2018 I DDC 133.4/40938 LC record available at https://lccn.loc.gov/2017046034

I Inscriptions,

Greek.

Contents

Preface

ix

Abbreviations for Corpora of Magical Texts Introduction

xiv

1

PART I.ARCHAEOLOGY Chapter 1. Distribution Chapter 2. Shapes

54

Chapter 3. Media

79

27

PART II. IMAGES Chapter 4. Action Figures

105

Chapter 5. Domestic Guardians

129

Chapter 6. Pharaonic and Ptolemaic Gods

PART III. TEXTS Chapter 7. Prayers

177

Chapter 8. Incantations

198

Chapter 9. Framing Speech Acts

221

Conclusions and Further Trajectories

238

148

viii

Contents

APPENDICES Appendix A. Summaries of Recipes for Protective Amulets Worn During Dangerous Rituals (from the longer PGM Handbooks)

Preface

263

Appendix B. Summaries of Recipes from a Curative Handbook Embedded in a Magical Handbook (PGM VII 193-214) 266 Appendix C. Summaries of Recipes from Smaller Fragments of Curative or Protective Handbooks 267 Appendix D. Summaries of Recipes from a Fragment of a Curative Handbook (Testamentof Solomon 18.15-40) 271 This study has a precursor of sorts, a book written more than a quarter century

Appendix E. Summary of Recipes from a Fragment of an Amulet Handbook (S&D 26-39) 274

ago about talismanic and apotropaic statues in ancient Greece and devoted mainly to the protective role of such images in the ancient Greek city. That book was well

Appendix F. Summary of Recipes Preserved by Marcellus of Bordeaux

276

received, but some reviewers were surprised that it did not contain a single illus-

Appendix G. Summary of Recipes Preserved by Alexander ofTralles

280

tration. And indeed, I must admit that at that time, because of my training and orientation as a philologist, I was focused almost exclusively on literary, rather than

Appendix H. Summary of Recipes Preserved by Aelius Promotus Appendix I. Summary of Recipes Preserved by Dioscorides

285

283

archaeological, evidence and that it did not occur to me to include any photographs or drawings. This volume, as the gentle reader will discover, has more than a hundred illustrations, a change that reflects both a great benefit and a great risk. Beneficial, I hope, is the way it brings together for the first time a wide array of

Notes

289

art-historical, archaeological, papyrological, epigraphical, and literary sources to

Glossary of Authors and Texts

409

bear on the subject of ancient Greek amulets; risky, because each of these scholarly traditions and each of the many subfields within them involve a bibliography

Glossary of Terms Bibliography

412

415

and an expertise that is beyond the human powers of a single scholar. There will, I am sure, be some details missed or evidence wrongly assessed, but my hope is that the experts, many of whom I thank below for their help, will overlook such lapses and errors and enjoy the wider and, I hope, richer panorama sketched in

INDICES General Index

these pages. This book has three thematic sections-"Archaeology," 457

Index Locorum

471

Ancient Words

483

Platesfollow page 224

"Texts"-that

"Images," and

could perhaps be read in isolation, the first by archaeologists, the

second by art historians, and the third by philologists, but I would strongly advise against proceeding in this manner, because such an approach would simply reify these traditional academic boundaries and undermine or at least ignore the central argument of this study that the popularity of easily identifiable amulets in the Roman period reflects a long process of accretion, whereby the traditional shapes and media surveyed in the first section begin in the imperial period to take on images and texts that tell us unequivocally and for the first time that these shapes and media were being used all along as amulets.

X

Preface

This volume began to take shape in Paris, when in the autumn of 2011 at the invitation of John Scheid, I gave four lectures at the College de France, each of which eventually became important parts of Chapters 1, 2, 4, and 8. I cannot thank John enough for the invitation, which helped focus what was at the time a somewhat

Preface

magic and even more so for his skepticism and goodwill in commenting on my work either from the audience or on the margins of the page. Over the past eight years, I have also visited a number of museums and libraries to photograph and study their magical gems, papyri, and other amulets, both

inchoate project, and I thank the audience for their questions and especially for their skepticism about some of the arguments that I eventually overhauled or abandoned.

for my own continuing education and for my contributions to the Campbell Bonner Database. At nearly every stop I encountered extremely dedicated and helpful

Roughly two years later I returned to Paris to take up a fellowship at the Institut

individuals. Early on, Matilde Avisseau-Broustet (Cabinet des Medailles), Yekat-

d'Etudes Avancees de Paris, where, thanks to the collegial atmosphere created by Gretty Mirdal and Simon Luck, I finished a full draft of all of the internal chap-

erina Barbash (Brooklyn Museum), Chris Entwistle (British Museum), Ken Lapatin (Getty Villa Museum), Andrew Meadows (American Numismatic Society), and

ters. Gratitude is also due to the University of Chicago's Paris Centre, especially

Brian Rose (University of Pennsylvania Museum of Archaeology and Anthropol-

to Robert Morrissey and Sebastien Greppo, for various kinds of additional support that made my year in Paris especially productive, and to Fran~ois Lissarrague, who

recent assistance, I am grateful to Giovanni Avagliano (Paestum Museum), Martina

kindly lent me his desk, computer, and scanner at the Gernet-Glotz Library. I had, in fact, been working on individual sections of this book as early as the autumn of

Bagnoli (Walters Art Museum), Lucilla Burn (Fitzwilliam Museum, University of Cambridge), Robert Daniel (Institut for Altertumskunde, Cologne), Susanne

2008, when, as a fellow at the Institute of Advanced Studies at Princeton, I wrote a long article on metrical incantations inscribed onto amulets that eventually be-

Ebbinghaus (Harvard Art Museums), Kay Ehling (Staatliche Mi.inzsammlung,

came a large part of Chapter 8. I am grateful to Heinrich Von Staden for conver-

Feldman (Department of Near Eastern Studies, University of California-Berkeley), Jasper Gaunt (Michael C. Carlos Museum, Emory University), Joe Greene (Harvard

sation and encouragement during my time at the Institute, as well as to the other fellows and especially to the librarians for their help. In the winter of 2009, as a research fellow at the Getty Villa Museum in Malibu and with the guidance of Ken Lapatin and Fran~ois Lissarrague, I began to make some first tentative steps toward thinking about the images on ancient Greek amulets and how one might go about interpreting them. It was there that I first presented the material that now makes up the second half of Chapter 4.

ogy) were exceedingly helpful and allowed extended or repeated visits. For more

Munich), Sonia Focke (Staatliches Museum Agyptischer Kunst, Munich), Marian

Semitic Museum), Juliet R. Graver Istrabadi (Eskenazi Museum of Art, Indiana University),Todd Hickey (Tebtunis Papyri, Bancroft Library, UniversityofCaliforniaBerkeley), Ulla Kasten (Babylonian Collection, Yale University), Christine Kondoleon (Museum of Fine Arts, Boston), Christopher Lightfoot (Metropolitan Museum of Art), Freiderike Naumann-Steckner (Romisch-Germanisches Museum, Cologne), Dirk Obbink (Oxyrhynchus Papyri, Ashmolean Museum),

Throughout the past eight years, I have been happily associated with the Campbell Bonner Magical Gems Database, which continues to flourish at the Musee des Beaux Arts in Budapest under the wisdom and energetic leadership of Arpad

J. Michael Padget (Princeton University Art Museum), Irini Papageorgiou (Benaki Museum, Athens), Dieter Quast (Romisch-Germanisches Zentralmuseum,

Nagy, whose keen eye and friendship have also been crucial to the evolution of the volume at hand. Attilio Mastrocinque, another dear friend and associate of the

Don C. Skemer (Rare Books Collection, Firestone Memorial Library, Princeton

Mainz), Stephen Quirke (Petrie Museum of Egyptian Archaeology, London), University), Rachael Sparks (Institute of Archaeology, University College London), Rudiger Splitter (Antikensammlung at the Museumslandschaft, Kassel),

Database, taught me how to photograph these magical gems during a marathon visit to the basement of the Kelsey Museum in Ann Arbor, Michigan. His exqui-

Jeffrey B. Wilcox (Museum of Art and Archaeology, University of Missouri),

sitely photographed and commentated corpora of the magical gems in the major Italian collections (2003 and 2008) and in the Cabinet des Medailles in Paris (2014)

and Terry Wilfong (Kelsey Museum of Archaeology, University of Michigan-Ann Arbor).

were timely and crucial research tools for this volume. I cannot thank him and

Finally, as my research for this volume often required me to venture into areas of study beyond my expertise, I am especially grateful to the following individuals

Arpad enough, as both read and commented on the entire manuscript and saved me from numerous errors. Other associates in the group include Simone Michel-

for answering questions, suggesting bibliography, reading parts of the manuscript,

von Dungern, author of the marvelous 2001 edition of the magical gems in the British Museum and an indispensable 2004 study of all the extant gems; Erika

or sending me their work in progress: Yekaterina Barbash, Caitlin Barret, Jacco Dieleman, Ian S. Moyer, Joachim Friedrich Quack, Robert Ritner, and Emily Tee-

Zwierlein-Diehl; Veronique Dasen; Jeffrey Spier; and Richard Gordon. To Richard I owe additional thanks for his numerous and penetrating studies on ancient

ter for guidance with the Egyptian material; Walter Farber and Joann Scurlock for the Mesopotamian; Brien Garnand, Carolina Lopez-Ruiz, and Phil Schmitz

xi

xii

Preface

Preface

for the Phoenician and Carthaginian; Bruce Lincoln for the Persian; Gideon

has become normal English usage (e.g., Socrates or Theophrastus) and in other

Bohak and Ortal-Paz Saar for the Jewish; Cliff Ando for the Roman; Veronique

cases to transliterate directly from the Greek (e.g., Dike or Ladike). In my trans-

Dasen, Patricia Gaillard-Seux, and Paul Keyser for the medical; Tom Carpenter,

lations of Greek and Latin texts, I use a very simplified system of brackets: square

Jas Elsner, Emanuel Meyer, Sarah Morris, Marina Piranomonte, Verity Platt, James

brackets [ ] indicate a lacuna in the ancient text that has been filled in by modern

Redfield, Joe Rife, Richard Veymiers, Paolo Vitellozzi, and Anastasia Zografou for the archaeological and art historical; Marianna Dagi, Robert Daniel, Todd

editors who extrapolate the missing word or words from the surrounding text, while parentheses ( ) are used to supplement the text by providing extra words or

Hickey, Raquel Martin Hernandez, Roberta Mazza, Dirk Obbink, and Sofia

phrases or, more rarely, giving in transliteration the original Greek word or words,

Toralles Tovar for the papyrological; and Alexander Hollmann and Roy Kotansky

with the goal of making a difficult passage less opaque. A glossary of perhaps un-

for the epigraphical. For those who I have unwittingly left off this long list, I offer my deepest apologies. I should close this list of acknowledgements by thanking

familiar authors and texts appears at the end of the volume, followed by another of perhaps unfamiliar technical terms. The nine appendices summarize a variety

again Roy Kotansky, who is in fact the amulet expert that I have known longest

of ancient formularies or handbooks that preserve recipes for protective and cura-

and with whom I co-wrote an article nearly thirty years ago about a gold-foil amulet in Connecticut: Roy, in addition to his friendship and his magisterial Greek

tive incantations or amulets.

Magical Amulets, which has been at my side throughout this project, has generously shared his notes, drawings, and insights; answered what must have seemed an unending series of email questions; and read and commented on the entire manuscript. In addition to the residential fellowships mentioned above, I am also deeply grateful for grants from the Loeb Foundation and the University of Chicago's Women's Board. Also in Chicago and for various kinds of editorial or technical help I am very thankful to Jeremy Brightbill, Kassandra Jackson, Thomas Keith, Megan Nutzman, and Walter Shandruk and for bibliographic genius to Cathy Mardikes at the Regenstein Library. Most especially, I thank my friend and former dean, Martha Roth, for her continual support of this project, both financial and otherwise. Finally, this work has derived continual energy, inspiration, and goodwill from my friends, co-teachers, and co-conspirators at the Center for the Study of Ancient Religions at the University of Chicago: Clifford Ando and Bruce Lincoln. I consider myself extremely lucky to have spent the past twenty-five years in such a rich and dynamic intellectual environment. Preliminary arguments for this volume appeared earlier as articles in Classical Antiquity; ClassicalPhilology;Greek, Roman and Byzantine Studies;Kernos,Journal of Roman Studies; MHNH· Revista internacionalde investigaci6nsabremagia y astrologiaantiguas;Museum Helveticum;Zeitschriftfur Papyrologieund Epigraphik;and a number of edited volumes. I have tried to make the book more user-friendly for nonspecialists by leaving out the original Greek texts and citing in transliteration, only when absolutely necessary, individual Greek words or phrases. I have, however, treated many of these texts in detail in previous publications, to which I direct the reader in the notes. In transliterating Greek names, it has seemed reasonable, if not entirely consistent, to use the familiar Latinized spelling for well-known names for which this

xiii

Abbreviations for Corpora of Magical Texts

LIM

A. Mastrocinque (ed.), Les intaillesmagiquesdu departement des Monnaies, Medailles et Antiques (Paris 2014).

Abbreviations for Corpora of Magical Texts

LIMC

Lexicon IconographicumMythologiae Classicae(Zurich 1981-present).

M&M

H. Philipp, Mira et Magica: Gemmen im AgyptischenMuseum der

Staatlichen Museen (Mainz am Rhein 1986). PDM

Demotic Egyptian Magical Texts translated in GMPT.

PG

J.-P. Migne, PatrologiaeCursus Completus, Series Graeca (Paris 1857-1936).

PCM ACM

M. Meyer and R. Smith, Ancient Christian Magic: Texts of Ritual

Powerfrom Ancient Egypt (San Francisco 1994). AGD

Antike Gemmen in deutschenSammlungen, 5 vols. (Munich 1968-75).

BGU

AegyptischeUrkunden aus den KoeniglichenlStaatlichen Museen zu Berlin. GriechischeUrkunden (Berlin 1895-2005).

EM

S. Michel, Die magischenGemmen im BritischenMuseum, 2 vols. (London 2001).

CIIP

H. Cotton, E. Lupu, and W Ameling (eds.) Corpus inscriptionum

Iudaeael Palaestinae:A Multi-Lingual Corpus of the Inscriptions from Alexander to Muhammad (Berlin 2010). D&D

A. Delatte and P. Derchain, Les intaillesmagiquesgreco-egyptiennes de la BibliothequeNationale (Paris 1964).

D&E

The Latin Lapidary of Damigeron-Evax cited according to the text of Halleaux and Schamp (2003) 230-90.

DMG

S. Michel, Die magischenGemmen: Eine Studie zu Zauberformeln und magischenBilderen auf geschnittenSteinen der Antike und Neuzeit (Geissen 1997).

GMA

R. Kotansky, GreekMagical Amulets, vol. 1, Papyrologica Coloniensia 22.1 (Opladen 1994).

GMPT

H. D. Betz (ed.), The GreekMagical Papyri in Translation (Chicago 1986).

Heim

R. Heim, Incantamenta Magica Graeca-Latina, Jahrbiicher for classische Philologie Suppl. 10 (Leipzig 1892).

Kerugma

The Greek lapidary OrpheosLithika Kerugma cited according to the text of Halleaux and Schamp (2003).

K. Preisendanz [and A. Henrichs], Papyri GraecaeMagicae: Die

GriechischenZauberpapyri2,2 vols. (Stuttgart 1973-74). S&D

The Greek Lapidary of Socrates and Dionysius cited according to the text of Halleaux and Schamp (2003) 166-77.

SEC

Syllogeinscriptionum Graecarum (Leipzig 1883-present).

sec

A. Mastrocinque (ed.), Syllogegemmarum gnosticarum, Bollettino di Numismatica Monografia 8.2.1 and 2 (Rome 2003 and 2008),

2 vols. SM

R. Daniel and F. Maltomini, Supplementum Magicum, 2 vols. Papyrologica Coloniensia 16.1 and 2 (Opladen 1990 and 1991).

SMA

C. Bonner, Studies in Magical Amulets Chiefly Graeco-Egyptian, University of Michigan Studies, Humanistic Series 4 (Ann Arbor 1950).

xv

Introduction

The Roman Empire, especially in the eastern half of the Mediterranean basin, was marked by an explosion of images and texts in a variety of media-metal, papyri, mosaics, gemstones-all designed to protect, heal, or grant some abstract benefit to the persons who wore them on their bodies or placed them in their homes. By much older scholarly accounts, this proliferation of readily identifiable amulets could best be explained by a sudden need for magic or by a precipitous rise in superstition or anxiety in this period, connected perhaps with the "internal breakdown of Greek rationalism," the "infection" from or "invasion" of peoples from the east bringing their "ridiculous superstitions," or the festering of a "rotting refuse heap" of pagan religions. 1 Such tendentious statements have, of course, disappeared nowadays, but their influence lingers, as in the overrepresentation of Egyptian designs in museum corpora of the "magical" gems; the overemphasis on prejudicial literary sources that depict Egyptian, Syrian, or Jewish magicians traveling throughout the empire peddling their wares; or the designation of some amulets as a "Greco-Egyptian" phenomenon-that is, produced at a time when and in an outlandish place where the Greeks were no longer truly Greek. 2 I will argue that these amulets were not invented in this period as a result of any alteration in the Zeitgeist or a tidal wave of "oriental" influence. Rather, they become visible in the archaeological record due to a number of technical innovations and transformations, the most important being the increased epigraphic habit of the imperial period, 3 the miniaturization of traditional domestic amulets like the triple-faced Hecate, or the utilization of newly crafted iconography, such as scenes of "gods in action" battling dangerous foes or outright inventions like Sarapis, a god created by the Macedonian king of Egypt. To put it another way, it is only when explicitly protective or curative texts or "weird" images unfamiliar to the classical Greek canon are added to Greek ornaments or doorways that modern observers realize that these objects were thought to have the power to protect or heal. The real

2

Introduction

Introduction

Figure 0.1 Thunderstone from Ephesus engraved with the Pantheos and magical names surrounded by an ouroboros (Royal Ontario Museum); drawing after Iliffe (1931). Figure 0.2 Thunderstone from Argos engraved with a gigantomachy below and a scene of Mithras and the bull surrounded by magical names; drawing after Cook (1925) fig. 390.

question, then, is not why we are able to identify so many amulets in the Roman imperial period, but rather why we fail to find them in the preceding centuries. 4 We can illustrate the problem clearly in the case of the so-called thunderstones: stone axe heads originally created and used as tools in Neolithic times, 5 that sometimes appear in much later archaeological sites or graves because they were thought to have some special power to protect buildings and people from lightning and fire.6 Our clearest literary evidence for the Greek practice, however, is

(perhaps Athena) is about to stab a tiny snake-footed "giant" with her spear,11 and

very late: Timotheus of Gaza, a fifth-century CE Greek author, tells us: "You will

the upper register shows a simplified version of the well-known Mithraic icon: the

have an amulet against a thunder blast, if you inscribe a thunderstone with the

god kneels on the back of the bull and stabs it, while three animals surround it

letters aphia aphrux and keep it in your house." 7 In the Roman period at least a

from below.12 And in Roman Pergamum, an enterprising artisan even sliced thun-

dozen of these axe heads were engraved with Greek letters, symbols, or images

derstones lengthwise-presumably

and were used as amulets, the larger ones presumably for homes and the smaller

faces with text and symbols typically found on amulets (Figure 0.3).13 Archaeolo-

for the body. For example, a thunderstone said to come from Ephesus 8 was incised

gists can, moreover, surmise that smaller prehistoric axe heads have been reused

in Roman times with the kinds of designs that appear frequently on the amulets

as amulets, when they are discovered in graves on the neck or chest of the de-

discussed in this volume (Figure 0.1): an image of the Egyptian "pantheistic de-

ceased or have been perforated for suspension. 14 In fact, scores of such repurposed

ity" (discussed in Section 5.5) and magical names, all encircled by a serpent eating

axe heads have turned up in graves from the Bronze Age to the Byzantine, 15 of

its own tail, a device that the Greeks called ouroboros("tail devourer"). 9 Another

which only thirteen, by my count, were incised with text or images. These en-

large thunderstone from Argos (Figure 0.2) was engraved with two different scenes

graved versions all date to the Roman period, use the Greek alphabet, and with

of "gods in action." 10 The bottom half depicts a gigantomachy, in which a goddess

one exception, probably come from eastern Greece. 16 Thus it seems that only the

to increase his profit-and

inscribed their

3

4

Introduction

Introduction

We shall see that the Greeks were already using amulets in Periclean Athens, some of which had indeed been borrowed from the east, 18 as well as others that were home grown. A second problem is the overabundance of Roman-period data from Upper Egypt and the scholarly habit of extrapolating from these data wider conclusions about the use of magic and the activities of professional magicians operating elsewhere in the empire. As we shall see, this overemphasis on the Egyptian contribution simply reifies older ideas and theories about the "oriental" origins and inherent strangeness of Greek amulets in the Roman period, and it allows scholars to ignore a number of other equally important kinds of evidence, at least in the case of amulets, for a more balanced understanding of a wider range of agents-such mothers-who

as jewelers, gem cutters, root cutters, homeowners, and even grandcreated amulets and who passed along traditional recipes both

orally and in written form. Finally, there is the wider prejudice that amulets are by definition "magic," when in fact they perfectly straddle the outdated divide between "magic" and "religion" by including both "magical" images (i.e., apotropaia) and texts (i.e., incantations) that aim at automatically driving away diseases and other evils, and "religious" images (i.e., of the gods) and texts (i.e., acclamations or prayers) to invoke or otherwise encourage the presence of beneficial and superhuman helpers to protect, rescue, or heal the person who wears the amulet.19 In this volume, I follow current scholarly practice in using the adjective Figure 0.3 Vertical slices of a thunderstone from Pergamum inscribed with magical names and symbols; drawing after Cook (1925) fig. 392.

"magical" in an extremely limited way to describe types of evidence-for example, "magical names" or "magical gems" in places where one might just as easily have used the adjective "nonsense" to describe such texts or "weird" to describe such gems.

Greeks in the eastern Mediterranean and only in imperial or late antique times felt compelled to enhance the apparently inherent protective power of these axe heads by adding texts, images, and symbols. Indeed, there are not to my knowl-

0.1 Scope

edge any extant examples of such thunderstones from Roman times inscribed with Latin, Aramaic, or demotic Egyptian. 17 This pattern cannot be explained by

Before we begin, it will be good to parse the title of this volume in order to fur-

some newly arrived superstition or anxiety, but rather by two important and specific changes in the Greek development of this tradition: (i) the rise in epigraphic habit of the eastern Roman Empire, which encouraged the inscription of incantations or prayers that had previously been recited over the stones; and (ii) the adaptation and miniaturization of powerful images-like

Pantheos, the gigantomachy, or

Mithras stabbing the bull. Our failure to recognize these earlier Greek amulets also stems to some degree from a number of scholarly prejudices or practices. First and foremost, perhaps, is the insistence-overt or not-upon the exceptionality of the classical Greeks in comparison to so many of their neighbors, for example, the Egyptians, the Assyrians, the Phoenicians, and the Etruscans, all of whom made extensive use of amulets.

ther establish its parameters. "Amulets" refers to any object-plant,

animal, or

mineral; natural or manmade, image or text-that the Greeks placed on their bodies, domestic animals, homes, ships, vineyards, or cities in the hope of protecting themselves, of curing some illness, or of gaining some benefit, usually understood in the abstract-for example, charisma, prosperity, or victory. 20 Body amulets used in healing were popular in all periods, and we can trace their importance through a series of passages that reveal both incantations and/or amulets as essential parts of traditional Greek medicine. The poet of the Odyssey,for example, tells us that Odysseus' cousins know how to bind a serious wound with a bandage and to stop the bleeding with an incantation (epoide),suggesting that both were part of the folk knowledge of Greek hunters, not that of professional

5

6

Introduction

healers.

21

Introduction

And in the Homeric Hymn to Demeter, the disguised goddess, when

angling for a job as a wet nurse in a royal house, claims to know "antidotes"presumably amulets or incantations-effective

in keeping disease, demons, or

witchcraft away from a newborn child. 22 In the late archaic period, Pindar describes in some detail how Asclepius, while still a mortal doctor, healed his patients: "And whosoever came to him affiicted by natural illnesses, or their limbs injured by grey bronze or stones far-flung, or with their bodies ravaged by summer's fever

"treated bites from asps and the clouding of the eyes, not with recognized drugs

(pharmaka), but rather with amulets (periapta)and incantations (epaismata),"and indeed Augustine, Caesarius of Aries, and Gregory of Tours all complain that their parishioners in North Africa and Gaul were using the same. 32 And when Gregory Nazianzus says that Christians had "no need of amulets (periammata) or incantations (epaismata),"his tenth-century commentator Basilius helpfully explains both terms in great detail: 33

or winter's chill, these he frees from the bonds of every sort of pain, tending some with gentle incantations (epoidai),giving others soothing herbs to drink or tying on (periapton) herbs to their limbs from every side, and still others he cures by incisions." 23 Here Pindar neatly summarizes the various curative techniques of

Amulets: the bits of colored thread round wrists, arms, and necks; and moon-shaped plates of gold, silver or cheaper material, which foolish old women fasten upon infants.

his day,24 and in this medical kit we find four procedures: incantations, medicines, amulets, and surgery. 25 Pindar, moreover, and all the authors who follow him use the same vocabulary to describe these amulets and incantations. For the former, they regularly use the Greek verb "to tie on" (periaptein)or the nouns related to it,

Incantations:chants sung over children by the same old women, muttering to avert evil, at the same time licking the babes' foreheads with their tongues, spitting and blowing to each side.

literally the "things tied on" or "amulets" (periammata and periapta),26 and for the latter, they employ the verb "to incant" or "chant over" (epaidein)or the nouns for "incantations" cognate to it (epoidaiand epaismata).27 In Plato's Charmides, Socrates even combines an amulet and oral charm in a

If we read only the philosophers and the bishops, we might indeed be convinced that by the Hellenistic period, the Greeks had completely dismissed both amulets and incantations as nonsense, but when Basilius twice mentions "old women" as

single cure when he describes a remedy for headache that he learned from a Thra-

the agents and children's bodies as the sites of these practices, we can see instead

cian healer: "[The remedy] was a leaf, and an incantation (epoide)was (i.e., sung)

that the use of these allegedly "irrational" practices was still deeply embedded in

over it, and if one sang the incantation and at the same time made use of the leaf,

family life despite the repeated calumnies of some elite males.

the cure would make the patient healthy in every way, but without the incanta-

It is important to stress, finally, that outright criticism of amulets seems to

tion, there was no profit from the leaf."28 In yet another passage, Socrates gives us

have been a late phenomenon. Pre- Hellenistic Greek authors who are hostile to

a list of curative techniques similar to Pindar's: until a man learns to behave well, he

other popular forms of healing do not attack the use of amulets, although we might

says, "neither medicines, nor cautery, nor surgery, nor incantations (epoidai),nor

have expected them to so. The Hippocratic On the SacredDisease,for example, a

amulets (periapta) will help him."

29

Here we perhaps see the first, albeit subtle,

fifth-century BCE diatribe against rival treatments of epilepsy and other spasmodic

steps taken in the world of philosophy to isolate what we modems might call

diseases, criticizes fumigations, purifications, and incantations, but makes no men-

"medical" techniques (the first three on the list: drugs, cautery, and surgery) from

tion of amulets, presumably because in this early period, amulets were not yet

the remaining "magical" ones (incantations and amulets). The pairing of incanta-

used to cure these dramatic types of ailments or because the makers of amulets

tions and amulets indeed becomes in later texts the standard way to summarize

were not yet perceived by the Hippocratics as serious competitors in the market-

"irrational" nonmedical procedures in order to criticize them. The Greek histo-

place of cures. And in the fourth century, Theophrastus in his satirical sketch "The

rian Polybius writing in the second century BCE, for example, mentions them

Superstitious Man" mentions purifications, omens, and other forms of ritual be-

both in a metaphor he uses to explain why in times of war, previously sensible men

havior, but neither incantations nor amulets.

make bad decisions: they are like sick people, he says, who, when the usual medi-

however, Plutarch in his own treatise On Superstitiondoes say in passing that when

cal procedures fail to work, turn instead to an incantation (epoide)or an amulet

a superstitious man was faced with personal illness, financial or political setbacks,

(periamma).30 This binary way of summarizing undesirable healing practices, often as a prelude for banning them, persists throughout later antiquity. 31 Julius Africanus, for example, an author who lived in the third century CE, is criticized for having

or the death of children, he would allow "old women, as Bion says, to bring what-

34

About four hundred years later,

ever they happened to find and tie it on (the verb is again periaptein),as if he were a peg."35 This series of passages, then, ranging from Pindar to the Christian apologists, suggests that from at least the late archaic period onward, the Greeks used

7

8

Introduction

Introduction

two important and sometimes interconnected modes of healing: charms sung over

exaggerated, but as we shall see, he does seem to have access to some genuine

the body or amulets tied on to it. The second word in the title of this volume-the

the common language and culture that begins to flourish in the seventh century

Persian lore about stones and amulets, lore that in turn originates much earlier in Mesopotamia. 43It is notable, however, that he makes no mention of any Egyptian role in either the invention of magic or its transmittal to the Greeks. 44A few gen-

BCE in mainland Greece and the Aegean littoral and eventually spreads by trade

erations later, however, Greek and Roman literary authors create a stereotype of

and colonization to the coastal areas of the Black Sea, Cyprus, Libya, southern Italy, Sicily, and even the mouth of the Rhone River and beyond. By the time these

the itinerant Egyptian magician or necromancer, whose appearance coincides with

adjective Greek-refers to

peoples were conquered by the Romans, Greek was the dominant language and culture throughout the eastern Mediterranean basin and even further east, thanks largely to the conquests of Alexander the Great. This now vast Greek-speaking area was, of course, still populated by non-Greek peoples-for example, Egyptians,

the sporadic appearances of Egyptian motifs on Greek curses in contemporary Athens, Carthage, and Rome. 45 And finally, in late antiquity, primarily Christian authors complain about the Jews as the foremost experts in magical healing, especially in exorcism, which in its classic form of adjuration does indeed seem to have been the invention of Hellenized Jews living in the eastern empire. 46 In

Syrians, Jews, Lycians, Lydians, Carthaginians, and others, many of whom naturally held onto their native languages and traditions-but as a result of unprecedented

successive waves, then, Greek and Roman authors living in the empire have com-

mobility and the vast growth of emigre populations in the major cities in the east,

of contemporary magic, and in each case these claims are greatly exaggerated or even mendacious, but they do contain kernels of truth.

Greek language and culture were adopted by many, at least those living in urban centers. 36

It is not at all surprising, therefore, that in the late Hellenistic and Roman periods, the Greek language became the most important vehicle for the production of textual amulets and for the collection and transmission of recipes for many other types of amulets-even in Egypt and Italy. 37The importance of the Greek language is illustrated by the fact that the extant amulets in the Roman period are only rarely inscribed with any of the other important languages of the empirefor example, Latin, Aramaic, demotic Egyptian, or Punic. 38 On rare bilingual amulets, moreover, although the name of the patient and goal of the amulet are occasionally expressed in the local tongue, the names of powerful gods and other forces are preserved in the original Greek, presumably copied from a Greek handbook. 39It is also notable that, although in the imperial period isolated verses of Homer are repeatedly quoted in Greek as protective incantations or inscribed on amulets, only a single Latin verse ofVergil seems to have been used in this capacity.40Finally, many of the common features of amulets-like the use of the seven vowels or palindromes-are of Greek origin, and even the outlandish symbols (charakteres),once thought to have evolved from Egyptian hieroglyphs, are now recognized primarily as distorted letters of the Greek alphabet. 41 As mentioned earlier, the Greeks clearly do borrow a number of curative speech acts and amulets from the east, some of them quite early on, but not solely, as it is sometimes assumed, in Roman times and not solely from the Egyptians and the Jews. This prehistory was not, in fact, unknown to the Greeks. Pliny the Elder, for example, depending heavily, as usual, on Hellenistic Greek sources, complains at length how the Persians invented and spread magic to the Greek world as early as the classical period. 42 His allegations are for the most part greatly

plained about the Persians, the Egyptians, and the Jews as the sources and agents

In the past scholars overemphasized the degree of Egyptian influence, in large part because they gave far too much weight to the so-called Theban Magical Library,47 a hoard of papyri rolls and codices from Upper Egypt thought to be the working handbooks of native Egyptian priests struggling in late antiquity to reconfigure their priestly and scribal selves as itinerant specialists who repackage traditional Egyptian magic for Greek and Roman consumption and who, through the process of stereotype appropriation, even assume the stereotyped roles found in Greek novels and satires. 48 This remains, perhaps, a persuasive hypothesis for the limited area of late antique Upper Egypt, 49 but the idea that these Theban handbooks primarily aim to present arcane Egyptian rites to foreigners living in Egypt can no longer be substantiated by the evidence of the recipes 50or the amulets themselves. 51This is especially true for the amulets, in part because there is very little overlap in content between the recipes in the Theban handbooks and the inscribed amulets from other parts of the Mediterranean world, and in part because the literary accounts of itinerant magicians do not describe them making amulets or singing curative incantations, but rather performing elaborate erotic curses or divination spells. The extent of Jewish influence on Greek magic has been similarly overstated. 52The frequent use of the names and epithets of Yahweh (Iao in Greek) and the angels is no longer understood as evidence of strong Jewish influence and authorship, and it now seems clear that many of the late antique and medieval recipes and amulets written in Aramaic and Hebrew actually derive from Greek handbooks and tradition, rather than the other way around. 53Scholars, in short, no longer support the older view that Greek magic was deeply influenced by Jewish or Egyptian (i.e., "oriental") practices, but in fact they now see significant Greek influence on magical practices in the east. The widespread use of the

9

10

Introduction

Introduction

exorcism formula is, of course, a special case, which as we shall see in Section 9.1,

above, been overly reliant on prejudicial Greek and Roman literary sources and

first appears in literary texts of the first two centuries CE, which report the exorcisms performed by Jews in Ephesus or "a Syrian from Palestine." The earliest extant

the idiosyncratic papyrus handbooks from Upper Egypt. The last four words in the title of this volume-"in Roman Imperial Times"-are far less controversial: they

exorcistic amulet is, however, inscribed in Greek, not Aramaic or Hebrew, and was found in Beirut. 54

mark this volume as a historical study of the transformations that occur during

In the specific case of amulets, Campbell Bonner, in an influential formula-

used in this volume, almost always refers to a time period rather than to a people

tion, argued that prior to the establishment of the Roman Empire there existed a number of"national traditions," about three of which we are well informed: Greek,

or culture. 59 This study is divided into three main sections. The first, "Archaeology," pro-

Egyptian, and Jewish. He argued, however, that in the early empire, an interna-

vides an eclectic survey from the late archaic period onward of the social range,

tional tradition of manufacturing amulets developed that freely chose from and often comingled these national traditions for clients of various ethnicities and re-

materials, and shapes of traditional Greek amulets, in order to provide a baseline of amulet use in the eastern Mediterranean prior to the establishment of the Ro-

ligions, a process helped by the growing use of Greek handbooks and the appear-

man Empire, although I do draw additional examples from later periods and from

55

first three centuries of the common era. As a consequence the word Roman, when

ance of itinerant scribes and magicians. Bonner's model has been a productive heuristic tool and will remain so, but it has two serious drawbacks. First, he based

the west in order to flesh out our picture of amulets that appear much earlier in the Greek world. In this first section I take special care to discuss artifacts from

his understanding of this international style on two ancient sources that greatly exaggerate the Egyptian elements: the late antique handbooks from the Theban

the entire Greek-speaking world, rather than to let the Athenian evidence domi-

Library discussed above and the modern category of magical gems that includes

examine vase paintings from classical Athens and terracotta votives from Cyprus, southern Italy, and elsewhere for images of the traditional amulets worn by an-

as its third feature "weird" images, mainly drawn from Egypt (more on this later). Second, Bonner generally ignored or downplayed the fact that the eastern Mediterranean was a place where such national traditions had been colliding and adapting for centuries prior to the arrival of the Romans-first in the Bronze Age, then again in the so-called orientalizing period and even in the classical period. Some Mesopotamian and Egyptian amulets were already being used by the Phoenicians or the Greeks long before Alexander conquered the east, and even more was accomplished under the Ptolemies. 56

nate the discussion, as is too often the case. In Chapter 1 ("Distribution"), we will

cient Greek children and women. Chapter 2 ("Shapes") surveys the use of Greek amulets, in many cases items, like the Gorgon's head, the crescent moon, or the phallus, that are widely recognized by scholars as amulets and that have forebears in the ancient Near East and Egypt. And in Chapter 3 ("Media") we turn to the various metals, stones, and other durable materials from which the Greeks in preRoman times manufactured amulets, such as gold, tin, hematite, red coral, and amber. We will also see how various handbooks, recipes, and literary sources can

But how, then, do we describe this complicated repertoire of magical amulets produced according to Greek handbooks and inscribed with Greek texts? Indeed,

alert us to an enormous number of perishable materials, such as cloth or leather containers, a variety of botanical amulets-especially roots and leaves-and soft-

if we decide to call (and many have done so) the Theban handbooks and most of the Greek magical papyri and gemstones "Greco-Egyptian," then we should prob-

tissue parts of animals as well, for example, eyes, tongues, hearts, and skins. As mentioned earlier, one of the most important transformations of amulets

ably label the large number of Greek curse tablets from Carthage "Greco-Punic" and those from Rome "Greco-Italic" and even describe the rite of exorcism as

in the Roman world is the addition of new images and text to previously unengraved

"Greco-Jewish." Such an approach could lead and in fact has led to some excellent

and uninscribed objects. In the second section of the book, "Images," we examine how the Greeks beginning in the Hellenistic period greatly enhance the reper-

fine-grained studies of epichoric ritual activity in which local geography trumps the common language and idiom. 57 In this study, however, I seek to describe and

toire of their amuletic images-mainly those of gods and heroes-that were thought to protect or heal. Chapter 4 (''Action Figures") focuses on the use of two-

trace the transformation of the common element in all of these hyphenated categories: the common "Greek" practice shared by all of the subgroups. The word

figured scenes in which powerful individuals triumph over dangerous animals or

Greek in my title is, then, a useful heuristic, albeit not always geographically or culturally accurate, category for discussing most of the amulets used in the eastern Mediterranean from at least the Hellenistic period down to the end of late antiquity. 58 It is offered here as a corrective to past studies that have, as I stressed

demons, such as Heracles strangling the lion or an equestrian hero stabbing a female demon. Such images provide a persuasive model for the desired cures: the suffocation of painful colic or the demise of a child-stealing demon. In Chapter 5 ("Domestic Guardians"), we turn to a series of pre-Roman protective statuettes from three different cultures-the

Greek herm, Roman Mercury, and Egyptian

11

12

Introduction

Introduction

Pantheos-and show how at least as early as the Hellenistic period such domestic images continue to protect and prosper houses, but are now also used in reduced size as body amulets, in the form either of miniature statues or of intaglios carved into the flat surfaces of semiprecious gems. And in Chapter 6 ("Pharaonic and Ptolemaic Gods"), I argue against the traditional fashion of talking generally about "Egypt" as the source and place of manufacture for Roman-period amulets and stress instead the crucial difference between "Pharaonic" (that is, pre-Hellenistic)

which in the case of Greek amulets would refer to the parade of nonsensical words and symbols or strange images that we find on so many Roman-period amulets. These names and images are striking and imaginative and give us real insight into the creativity and worldview of the professional magicians who created and adapted them. 60 In this study, however, I chose to downplay the nonsensical and the weird in order to stress instead the broad structural similarities between various kinds of amulets and thereby to assess the wider historical changes that occurred

images, often depicted in stiff poses with animal heads, which in the Roman pe-

during the Roman period in the Greek-speaking eastern Mediterranean. Thus,

riod are usually not popular on amulets outside of Egypt, and newly invented "Ptolemaic" or "Hellenistic" ones, such as Sarapis, Isis-Fortuna, and even Aphrodite Anadyomene.

for example, in the discussion of the transformation of spoken prayer into textual amulet in Chapter 7, we will not concern ourselves with identifying precisely which all-powerful solar god is invoked (Zeus, Helios, Re, Yahweh, Mithras, or

The third and last section, "Texts," examines how speech acts performed orally

some other entity with a long nonsensical name). Often many different names

in pre- Roman times to protect, heal, or gain abstract gifts are first inscribed on

(comprehensible and not) are invoked in the same prayer or exorcism, a practice

papyri and metal in the Roman period. We begin in Chapter 7 ("Prayers") with the use of vowels, divine names, acclamations, and prayers to invoke the presence

that suggests that in the Roman period, at least, some believed that such gods were interchangeable or that all nations worshipped the same sun god under dif-

of superhuman helpers, and we then go on to see how a traditional prayer to the

ferent names. 61 We might take as an example a magical recipe for a jasper amulet engraved with the sun and moon and the names Abrasax, Ia6, and Saba6th and

Samothracian gods for safe passage through storms at sea is in Roman times inscribed onto amulets used to ward off hail from vineyards and (by analogy) to avert a variety of cloudy or windy ailments from mortal bodies, for example, ophthalmia or headache. In Chapter 8 ("Incantations"), I discuss the evidence from pre- Roman Israel and Greece for famous singers, like David, Orpheus, or Empedocles, who were thought to sing incantations directly against demons and/or diseases. I then discuss a handful of early Greek experiments in inscribing lead and gold amulets and close the chapter by tracing the development of two popular types of Greek incantations from speech act to inscribed amulet: hexametrical verses that compel predatory animals and diseases to depart and iambic charms that

consecrated with the following prayer: "Complete this operation for me so that I may wear this power in every place, in every danger, without being smitten or afflicted .... I call upon you as 'Phno-eai-Iabok' according to the Egyptians, as 'Ad6nai Saba6th' according to the Jews, as 'King of all, ruling alone' according to the Greeks, and as 'Ouerto' according to the Parthians!" 62 These kinds of equations are not, however, a special feature of "magical" prayers, but typical of the times, as we can see, for example, in a prose hymn composed by the third-century CE writer Menander Rhetor as a model to be imitated by others: "Every city, land and nation do you control ... as 'Mithras' the Persians address you, as 'Horns' the The Chaldeans address you as 'Ruler of the Stars.'"63

stop the forward motion of venomous animals as well as the pathological flow of liquids like bile or rheum. This final section of the volume ends with Chapter 9

Egyptians ....

("Framing Speech Acts"), which discusses two other important innovations of the Roman period: the use of elaborate framing devices to embed formerly simple

on amulets as Artemis, Selene, Isis, Hecate, Persephone, and Ereshkigal. 64 The presence on this list of the underworld deities, like the Greek Persephone and Mes-

prayers and incantations in complicated hierarchies or divine histories. The first

opotamian Ereshkigal, might surprise some, but in fact, because in the Egyptian

is a relatively new and widespread phenomenon called exorcism, in which a demon is driven out of a host body by fumigations or verbal threats. The second is the far

and Mesopotamian systems both solar and lunar deities spend part of each day under the earth, they are also treated as powerful entities of the underworld, in such

less popular adaptation of short narratives (historiolae)of divine miracles or cures,

a way that the usual Greek distinction between the heavenly and chthonian realms is all but effaced and we have the odd situation (for the Greeks, at least) in which

which provide the model for the desired cure for the patient. Like the prayers and incantations, these framing devices were originally oral phenomena, but by the

We see similar strategies in the case of the lunar goddesses, variously invoked

solar or lunar gods can be invoked for necromancy. 65 We find a similar kind of

Roman period, Greeks begin to inscribe these as well onto papyrus, metal tablets, and gems.

internationalization in the treatment of divine images, whereby the clearly identifiable image of one god can be invoked by the name of another, as seen on a

One important aspect of ancient Greek amulets might have also been treated more extensively in this volume: Malinowski's famous "coefficient of weirdness,"

gem engraved on the front with the famous Pheidian image of the seated Zeus at Olympia and on the back with a short prayer: "Yahweh, save me!" or another

13

14

Introduction

Introduction

available from the Greek pantheon, such as the previously mentioned triple-headed Hecate (Figure 0.4). But much older images also persist unchanged from each of Bonner's "national" traditions-for

example, the mummiform

Osiris from

Pharaonic Egypt, Heracles and the Nemean lion from archaic Greece, or Mercury and his money sack from pre-imperial Italy.

0.2 Sources One challenge to writing such a monograph is the great number and variety of primary sources, which can be divided roughly into two categories. First are the amulets themselves, both uninscribed (for example, crescent moons or animal teeth), and those composed entirely of text (for example, small scrolls of papyrus or metal foil). Second and in some ways more important, because scholars have generally ignored them, are the many recipes for amulets, which tell us why and how they should be made and worn. These recipes date mainly to the late Roman and early Byzantine periods and survive in two forms, directly in the papyrus handbooks found in Egypt or indirectly in other kinds of collections, such as lapidary or medical treatises, or in more erudite encyclopedias, which survive mainly in Greek manuscripts compiled in Byzantine times, but which clearly preserve much earlier material. Pliny, of course, wrote in Latin, but since his chief goal seems to have been to preserve the knowledge of the past collected by earlier Greek authors, Figure 0.4 Greenish-brown jasper engraved with triple Hecate on the obverse and the name Ereshkigalon the reverse (BM 66); author's photographs used with the permission of the Trustees of the British Museum.

his Natural History often supplies important details about traditional amuletsmineral, floral, and faunal-known to earlier generations of Greeks.69 Other important sources for recipes or descriptions of amulets have already been touched upon: a wide range of ancient social critics, such as philosophers, medical writers, comic poets or satirists, and public officials of varying stripes, including Roman jurists,

with triple-faced Hecate on one side and the name "Ereshkigal" on the other (Figure 0.4).66 But we will also encounter completely novel and often hybrid im-

Talmudic rabbis, and Christian bishops, who beginning in the late classical period sometimes tell us what these amulets looked like and how they were worn. We

ages, the best known perhaps being a rooster-headed, snake-footed individual (Plate 1) whom scholars have named the anguipede (lit. "snake-foot"), an image that seems to combine the Egyptian tradition of animal-headed gods who hold

have already observed the prejudices and limitations of some members of this

flails, the Greek penchant for snake-legged giants, and the Roman cuirass and

sarcasm and exaggeration often supply us with the most important and-in the case of Aristophanes-quite early details that corroborate the archaeological and

group as sources for the study of incantations and amulets, and we can likewise anticipate a similar degree of distortion in comic authors, who despite their

shield-a shield whose concave interior is inscribed in Greek with the name of Yahweh.67 Another example might be the image of the Chnoubis serpent, which

visual evidence.70

combines the radiant head of a lion and the body of a serpent; in Plate 2 it appears on the opposite side of a gem engraved with Asclepius, the Greek god of healing,

Each corpus of extant amulets comes with its own distortions or limitations, and scholars who focus solely on a single source run the risk of the blind men in

and his wife Hygeia ("Health"). 68 The gem cutters do not, however, usually render female deities as hybrid creatures in such manner, although there is a clear ten-

the old story who each take up a single part of an elephant-the trunk, the tusk, or the tail-and end up with widely divergent ideas about the nature of the beast

dency on some amulets, as we shall see, to choose the weirdest traditional images

at hand. If we turn, for instance, to the papyrus amulets, which begin to appear

15

16

Introduction

Introduction

as early as the first century CE in Egypt, we find an emphasis on regional con-

complicated by the fact that ancient authors and modern scholars often disagree

cerns, as seen in the numerous amulets against malarial fever or scorpion sting. 71 But if we, instead, examine only the inscribed metal amulets found all over the

about or unwittingly confuse the nomenclature for the types of stones and even

eastern Mediterranean (but rarely in Egypt), we might understandably come to

their colors.79 Handbooks and recipes are our second primary source for ancient Greek am-

the conclusion that fever and scorpions were of little concern amid a much wider

ulets. They help us in many ways to imagine the full range of amulets used by the

array of diseases and dangers, especially demons and sorcery. And the Greeks used

Greeks, especially those materials that hardly ever survive antiquity, such as the

gemstones to cure still other complaints, especially gynecological or gastrointestinal ones. In short, none of these sources give us the full picture. Papyri amulets, because

ephemeral parts of plants or animals, as well as texts and images drawn on less durable media like leather, linen, or even leaves. Despite the late date of their com-

they rarely survive outside of Egypt, reveal the fears and concerns of persons in-

pilation or inscription, many of these handbooks preserve descriptions of amulets

habiting that distant corner of the Greek-speaking world, and we should not assume them to be typical of amulet production throughout the empire. The metal

and other kinds of information that go back at least to the Hellenistic period, if

amulets and the gems, on the other hand, give us a much wider perspective geo-

often collect and replicate amulet recipes from earlier handbooks, whose contours can occasionally be intuited in the surviving materials. so These recipes fall into

graphically, but a narrower one in terms of economic or social class, because they tend to be inscribed on valuable media, such as gold, silver, or semiprecious stones,

not earlier. The late antique Greek magical handbooks from Egypt, for example,

two categories. First are those designed primarily to protect a solo practitioner

and thus probably inform us only about those at the top of the socioeconomic ladder.

while performing potentially dangerous rituals used to summon powerful gods, demons, and ghosts (see Appendix A). Seventeen of the twenty-three examples

The gems are the largest and perhaps the most important corpus of extant amulets from the Roman period, but they are also the most difficult to interpret.

are found in the Theban handbooks discussed earlier and, not surprisingly, they

They number in the thousands if we take into account those in private collections,72 and they attract the most attention, perhaps, because they are sometimes engraved, as

include the highest concentration of Egyptian features, such as the udjat-eye of Horns or the scarab ring, and they do indeed give us a good idea of what Greek amulets may have looked like in a far-off corner of Upper Egypt in late antiquity. 81

we have seen, with strange and startling images (Figure 0.4, Plates 1 and 2). Archaeologists distinguish such "magical gems" from ordinary intaglios by three formal

The second type of recipe generally shows up in short extracts from handbooks concerned primarily with healing, for example, a contiguous sequence of short cu-

features: 73 (i) they are generally engraved on both sides, rather than one; (ii) they

rative recipes in PGM VII, each introduced by rubrics that specify their use (see Appendix B). Given the uniformity of these rubrics-all are indexed by disease in

are usually inscribed with Greek text that is not retrograde (that is, the gemstone was not used as a signet ring to seal documents); and (iii) they often carry non74

standard (that is, weird) Greek words, symbols, or images. These magical gems, moreover, tend to have flat surfaces and are cut from different and usually opaque

the same brief fashion (i.e., "For scorpion sting" or "For discharge of the eyes")and given the repetition of recipes for the same problem (for example, in the case

media in contrast to the convex translucent gems favored in the earlier periods. 75

of cough or migraine), it is probable that this series has been excerpted from an older collection of healing incantations and amulets, which seems to have circu-

Since the great majority of these gems were either surface finds or purchased on the antiquities market, we almost never know exactly where they were found and

lated in different versions. 82 These ten short recipes are also arranged in a special order that we sometimes find in Greek medical handbooks; they begin with cures

in what context. 76 And when scholars assert that the majority of these gems are "Greco-Egyptian" or that the majority of them were manufactured in Alexandria,

for the head and then descend from place to place down the body.83 Series like

they are confusing or combining the categories of "Pharaonic" and "Ptolemaic"

these give us a helpful glimpse of the range of simple and inexpensive cures someone might use at home to cure scorpion bite, headaches, coughs, or fevers.84 Three

(see Chapter 6), and they are ignoring evidence for gem production outside of Egypt, such as a recent study that shows that some hematite "magical gems" were

(nos. 3, 8, and 9) require verbal recitation while performing some kind of ritual with oil or knots, while the remaining seven call for the inscription of Greek

being manufactured in the Levant and in simpler forms that eschew Egyptian additions. 77 Beginning with Bonner much valuable work has been done, of

letters or symbols on papyrus, parchment, linen, or a leaf. The absence of gold

course, to establish standard types within the corpus-that is, cases where we can see a strong correlation between at least four features: type or color of gem, image, text, and the use to which it was put.

78

This work is useful, but it is often

or silver suggests a more modest socioeconomic context, and the divinities invoked clearly reflect the full range of Bonner's national styles, but without any strong Egyptian presence: Re, Zeus, Abrasax, Saba6th, and Castor, although the "Castor" here may not, in fact, be one of the Dioscuri. The word appears in a

17

18

Introduction

short incantation ("Castor Thab Thab") used in a cure "for swollen testicles," and given the popular ancient beliefs about the beaver (castorin Latin) and its testicles, 85 this recipe may preserve an old Italic prescription, not at all unlike the one preserved in the mid-second-century BCE Latin agricultural handbook of Cato the Elder, who prescribed for a dislocated bone that we manipulate a split reed while chanting motas uaeta daries dardaresastatariesdissunapitercontinuously. 86One need not, in short, look only to the mysterious east for incomprehensible incantations. Nearly a dozen other small papyrus fragments of such curative handbooks show the same pattern of rubrics keyed to diseases and the bundling of recipes for the same disease. 87We also see recipes for the same disease or the same part of the body in much smaller, nontraditional formularies, such as an ostracon "handbook" with only two recipes (separated by a horizontal stroke) for curing the poison of a scorpion sting or a small agate sphere that preserves (in a sequence again borrowed from medical treatises) a series of curative magical names or symbols, each designed to heal a different part of the head. 88These smaller "handbooks" may have been discrete "chapters" within a larger collection of recipes, for example, a pile of ostraca in a drawer, each inscribed with recipes for a different disease or a bag of spherical stones, each with cures for a different part of the body.89A collection of recipes for protective amulets incised on a copper-alloy plate from Sicily gives us important insight to yet another local tradition heavily dependent on the Jewish tradition about the amulets that Moses wore when he met face to face with Yahweh.90 Some papyrus fragments of earlier date and from areas of Egypt closer to

Introduction

Equally important sources for amulet recipes are the Greek lapidaries, which first appear in the Hellenistic period and probably draw heavily upon much earlier Egyptian and Mesopotamian sources. 96 The extant Greek texts usually take the form of encyclopedias: each entry begins with the name of a stone and a description of its physical properties, before finally listing its various uses as a medicine (if ground up) or as an amulet. Pliny is our earliest source for snippets culled from these earlier Greek writers, who range from the third-century BCE writer Sotacus97 to Pliny's contemporary Xenocrates of Ephesus, some of whose fragments are only preserved in later Arabic compilations. 98The fragments of the latter are especially important because they confirm the suspicion of many scholars that much of the gem lore reported without citation by Pliny also comes from Greek sources: for example, both Pliny and the later Arabic quotations of Xenocrates give similar accounts of the "eagle stone" that was tied to pregnant women to prevent a premature birth and otherwise to facilitate a timely one.99 Contemporary with the Hellenistic author Sotacus, on the other hand, are the equally lost and usually pseudepigraphic works of the "learned magicians" of the east, one, for example, who called himself "Ostanes," and another who pretended to be the Persian religious leader "Zoroaster" and recommended the power of the eagle stone and other amulets for epilepsy and hemorrhoids. 100A third author claimed to be the late classical Greek philosopher "Democritus," and he seems to have been, in fact, an Egyptian priest named Bolus of Mendes, who was educated in Greek and wrote in the first half of the second century BCE. 101Yet another

the Mediterranean likewise reveal a richer array of patently Greek traditions. For example, a small fragment of a first-century BCE handbook (the so-called Phil-

Hellenistic author masqueraded as the last Egyptian pharaoh "Nechepso" and wrote

inna Papyrus) preserves two hexametrical charms, one for headache and the other

amulet, and a defective copy of which Thessalus of Tralles claims to have studied in Alexandria in the time of Trajan. 102Pliny also cites two "Chaldaean" authors

for some kind of inflammation. Arranged like an anthology of poems, each incantation is introduced with the name of its female author and her home: "Philinna of Thessaly" and "[ .... ]a, the Syrian woman from Gadara."91 Another

a treatise on curative gems, which Galen quotes in his discussion of a digestive

working in the courts of eastern kings, Sudines, who was perhaps a seer for Atta-

important excerpt of a more practical handbook is embedded in Book 18 of a Greek

lus I (241-197 BCE), and Zachalias of Babylon, who is said to have lived during the reign of Mithridates IV (120-63 BCE).103 Pliny notes with disapproval a book

text known as the Testament of Solomon; this series of twenty-six recipes seems to

on stones written by this Zachalias, which claimed that jasper was beneficial if

have circulated independently, as early as the first century CE, as an independent handbook for curative amulets and incantations. 92In addition to treating fevers, chills, and other bodily infirmities, 93this handbook also deals with social or psy-

worn by litigants or petitioners appearing before a king. 104 He repeats, as well, the claims of the "Magi" that a special kind of"air-colored" jasper was "useful for those

chological ills, some of which can be remedied by placing amulets or texts on the house door.94We see, however, the same limited range of techniques: of the twentysix cures, fourteen are written on amulets or house doors, eleven are spoken over the patient-into his ear or over the wine or oil used in the cure-and a single recipe recommends the use of an image. 95And again the absence of valuable media like gold for textual supports suggests a lower socioeconomic bracket; there is one tin amulet, four papyrus, and the rest are unspecified.

who harangue the assembly," an amulet that also reflects a much older Babylonian tradition. 105 Pliny seems, in fact, to accept the general principle that Magian amulets can cure disease, but criticizes the idea that they can also help their owner persuade judges, kings, and popular assemblies or can protect them from bodily harm in battle. 106 Each of these Hellenistic authors can be traced by name to only a scant handful of fragments, but there is good reason to think that the tradition in which they worked was carried on in the Roman period. When Pliny, for example, reports a

19

20

Introduction

Introduction

further Magian claim that an amethyst will protect against spells if inscribed with the names of the sun and moon, or that smaragdus(a green stone conven-

are almost all in the section on onyx:113 Isis (no. 1), lion-headed Chnoubis (nos. 7, 4 Moreover, the references in 10, and 11), a scarab (no. 9), and Aphrodite (no. 12).11

tionally translated as "emerald") will provide similar benefits if engraved with an

this handbook to a reddish "rare stone" worn by the Persian king (no. 3) or the

107

eagle or a scarab beetle, parallels from the Roman world, but not before, suggest that Pliny perhaps encountered these "Magi" not in books, but rather as local

"Babylonian stone" worn by noblemen in the palace (no. 5) may reflect authentic

108

istic treatises discussed above.115 When compared with the amulet handbook em-

But the Hellenistic tradition of lapidaries continued as well, in a series Greek and

bedded in PGM VII, this unique sequence of recipes preserved by Socrates and Dionysius differs in two important ways: (i) only two of the fourteen stones have

magicians (magi, not Magi) selling such gems and boasting about their virtues.

then Latin texts assembled during the late Roman and Byzantine periods. The most important surviving group are closely related and seem to have evolved as 109

Near Eastern traditions, like those that Pliny and others found in the lost Hellen-

magical inscriptions, and as a consequence those that depict the Greek gods are to

(i) the OrphicLithica composed in the first half of the second century

us indistinguishable from many of the gems that our museums treat as mere or-

CE that takes its information from an earlier source and interlards it with long

naments or jewelry; 116 and (ii) the rubrics in the lapidary refer to media alone and thus would be of little practical use for someone looking for a cure (for example,

follows:

mythological digressions on the origins of each stone; (ii) the prose Kerygma, a Byzantine Greek epitome of the first twenty-five chapters of the Orphic Lithica,

for headache or fever), but would be helpful perhaps to gem cutters who were col-

which jettisons these digressions and adds important new material from earlier sources; and (iii) the continuation of the Kerygma (chapters 26-53) by a Greek

lecting or buying materials for amulets. Medical and pharmaceutical writers, both those who are favorable to the prac-

text attributed to Socrates and Dionysius, which seems, however, to have been

tice and those who are not, provide a third important source for amulet recipes.

penned much earlier in imperial times in the Greek east, perhaps Egypt. This

Some of these authors seem to be reading the same Hellenistic stone books as

last text is extremely valuable for the study at hand, because it begins with what appears to be a continuous section of entries preserved from an even earlier and

the lapidary authors. The Greek physician Galen (fl. 2nd century CE), for instance, wrote a treatise on the medical use of minerals in which he mentions that the

more practical handbook (Appendix E). There are fourteen gems in all: eleven carved with images, two inscribed with text, and one that "works well even

"foaming-moon stone" is useful for healing epileptics, but says that he has not

without any carving." 110 This last comment suggests that the author of this sec-

tried it on any patients, adding that "there are also some other stones that are tied on for such reasons, some that even have certain symbols (charakteres)and letters

tion, at least, expected most amulets to have images. And since elsewhere the extant lapidaries only rarely recommend images for amulets 111-an important

engraved upon them, just as the falcon-stone is also suitable against hemorrhoids, of which we, too, have made trial." 117 There is little doubt that Galen did test the

point to which we will return-it

appears that these fourteen recipes were ex-

efficacy of such stone amulets, because a little earlier in the same book, he de-

cerpted as a group from a different kind of handbook, one that specialized in powerful stones with images and one that was probably arranged by the color and

scribes an experiment that convinced him of the power of green jasper in curing digestive problemsY 8 Galen, in short, consulted earlier Greek treatises that included

type of stone: greenish stones first, reddish next, followed by generally whitish ones said to be onyx.uz The fourfold repetition of one rubric (''Another onyx")

curative stones, some of which were engraved with images or symbols. He says in

appears nowhere else in the lapidaries and is, as suggested earlier, a clear marker of a formal handbook. In this same embedded handbook, we once again see a mixture of Bonner's "national traditions," but in this case the images are drawn mainly from the classical Greek repertoire-for

example, Poseidon with his trident and a dolphin (no. 1),

Artemis with her hound (5), Athena with her helmet (6), and Apollo and Artemis together (8)-gods who are accordingly chosen for their traditional powers or spheres of influence: Poseidon protects at sea, Athena provides victory over enemies, and the presence of the divine twins Apollo and Artemis, themselves the successful product of a difficult birth, ensures the successful birth of the woman wearing the gem. Some Ptolemaic images cluster near the end of the list-they

a lost work, moreover, that although he originally thought that healing incantations were useless, "like old wives' tales," he came to realize that "there was power in them" after he saw that they cured patients stung by a scorpion or with bones stuck in their throat. 119 His critical engagement with curative incantations and amulets suggests that he took such techniques seriously and attempted to ferret out the effective from the bogus by experimentation. 120 Botanical amulets get a similarly mixed reception by ancient pharmacists. Already in the third century BCE Theophrastus disapproves: "What is said about amulets (periapta)and in general about aversion charms (alexipharmaka)for the body and the house is somewhat foolish and incredible! Thus they say that tripolion, according to Hesiod and Musaeus, is useful for every good purpose ... [and] that the plant called snapdragon produces fair-fame .... The man who anoints himself with this, they say,

21

22

Introduction

Introduction

wins fair-fame and they say the same result follows, if he crowns himself with the flower of the gold-flower plant." 121

of Amida, Marcellus of Bordeaux, and Alexander of Tralles. 128There are, how-

Despite his negative attitude, Theophrastus provides us here and elsewhere with important details about popular plant amulets and incantations, much of it gathered from interviews with contemporary "root cutters" and "drug sellers."122His

Trajan, prescribes numerous stones, flora, and fauna to be used as amulets, but no oral or inscribed charms, 129whereas the later authors include a wide array of incantations and amulets. 130In some cases the doctors even admit that these amulets

report that some lost poems attributed to Hesiod and Musaeus recommended the

supplement or even replace the medical treatment of difficult diseases. In a telling

use of plants as amulets, suggests a much older Greek tradition that is also hinted

confession, Alexander says that he appends recipes for amulets to his chapter on

at in the Homeric episode in which Hermes gives Odysseus a plant called moly to protect him from Circe's witchcraft. 123Two first-century CE writers, Dioscorides

gout "because some people, who are unable to persevere in their diet or to take their

and Pliny, preserve additional information. It has been pointed out how both writ-

for gout," and Pliny says that because ordinary medicines are practically useless

ers often distance themselves from these botanical amulets by saying, for exam-

against malarial fever, he will include a series of other remedies, including amu-

ple, "some people say" or "it is reported," but it is not stressed enough that such

lets.131In so doing, both authors echo, of course, the elite view discussed earlier

expressions also reveal that these recipes were probably traditional in their own day and thus predate the Roman period. 124And when Galen, on his part, criti-

that amulets are the last resort of patients after medicine has failed. But by sup-

cizes the botanical work of the polymath Pamphilus (ea. 60-80 CE) because he was distracted by silly incantations (used while cutting plants), 125his double in-

ever, important differences: Aelius, for example, a doctor alive during the reign of

medicine, force us to make use of (i.e., nonmedical) remedies and amulets (periapta)

plying these recipes, both authors also reveal their tacit approval. Soranus, a doctor in the generation that preceded Galen, may explain one possible motive when, in reference to a popular amulet for uterine bleeding, he says that such amulets,

dictment lets us know that some of the earlier Greek botanicals contained similar details.

although ineffective, should be allowed because they make the patient more hope-

The discussion of plants as curative amulets was complicated, however, by their

deed, when in the mid-fourth century CE Ammianus claims in passing that uttering

parallel use as internal medicines. Distinguishing the two became a touchstone by which Augustine in a later period attempted to outlaw the use of plants as am-

an incantation over a malarial patient was something "that even the medical authorities approved," we should hesitate before disbelieving him. 133The full range

ulets among his Christian followers: "It is one thing to claim that you will not

of amulet recipes preserved by the doctors of the Roman and late antique periods

have a stomachache, if you drink a certain herb after it has been ground up, but it

is underappreciated by scholars of ancient magic, who are often only looking for

is another thing to say that, if you wear that herb on your neck, you will not have

parallels for single incantations or images that appear elsewhere on extant amulets.

a stomachache. For in the first case there is a healthy concoction to approve, but in the later a superstitious sign to condemn." 126Galen, writing a few centuries ear-

scant handful have been discussed in detail, such as an inscribed amulet for eye

lier, took a somewhat different view. We saw how his experiments with some stone

ful.132Like Galen, later doctors may have also condoned curative incantations. In-

Of the thirty-seven amulet recipes collected by Marcellus, for example, only a

amulets convinced him of their effectiveness, and he came to similar conclusions

disease that quotes a verse of Homer and a pair of recipes that recommend using the image or symbol of Chnoubis on an amulet.134 And although Alexander of

about plants. For instance, when discussing the amulet recipes of the late first-

Tralles records eight different colic amulets, only two are regularly discussed: a

century CE medical writer Archigenes of Apamea (95-115 CE), who recommended

ring-stone depicting Heracles and the lion, and an iron ring inscribed with a hexametrical incantation. 135

placing a garland of polygonumon the head to cure headaches and a peony root on the neck for epilepsy, Galen describes how he experimented with the root by placing it on and then off the neck of an epileptic baby; he concluded in a quasiatomistic fashion that the amulet worked because particles of the root were inhaled by the child and thus prevented

0.3 Summary

seizures. 127

Galen is sometimes seen as a key protagonist in the triumphant evolution of

We have seen how the ancient Greeks as early as the classical period utilized, rec-

rational Greek medicine to the exclusion of "irrational" things like amulets and

ommended, and criticized the protective and curative powers of incantations, as

incantations, but this narrative elides Galen's experiments with both and ignores

well as those of various amulets, and how by the Hellenistic period there had de-

the many recipes for amulets still found in the medical books of other Roman-era

veloped a lively tradition of lapidary, medical, pharmaceutical, and other treatises

doctors, such as Aelius Promotus, as well as in the much later treatises of Aetius

or handbooks that collected, tested, and transmitted this lore for centuries to come.

23

24

Introduction

Some of this information, as in the case of the lapidary handbooks, seems to come from Mesopotamia perhaps by Persian intermediaries, while in other cases the sources may be native Greek root cutters. There is no doubt that the Greeks

PART I

were using many amulets before the arrival of the Romans, but the question remains why we have so little archaeological evidence for their existence. In the case of the botanical amulets, explaining this absence is quite easy, because such materials would rarely survive to be recovered by archaeologists. In the case of stones, however, we have already suggested one major problem in the evidence: the lapidaries and indeed all our handbook sources only rarely recommend engraving images on amuletic stones, and even when they do suggest images of Poseidon or Athena, only once do they tell us to add an inscription that might call our attention to their use as amulets.

Archaeology

CHAPTER 1

Distribution

Pre-Roman Greek sources provide, then, a number of descriptions of"amulet" and "incantation" as generic categories, but hard literary evidence for their specific forms or social contexts is difficult to come by. One tantalizing datum is a passing reference to an amulet that the famous Athenian statesman Pericles wore when he was dying of the plague in the second half of the fifth century BCE: "Theophrastus in his Ethics, while discussing whether one's character may be changed with one's fortune ... , relates that when Pericles was ill, he showed to one of his friends, who had come to visit him, an amulet (periapton)which had been attached around his neck by his women, as much as to say that he must be very sick indeed to put up with such folly (abelteria)as that." 1 In this passage we do learn something about the domestic context of amulet making and the role that women played in it, but unfortunately Theophrastus does not give us any hint as to the shape or medium of this amulet, information that might help us to identify similar amulets in the graves of Pericles' generation, many of whom died of the plague and some of whom presumably donned curative amulets like his after falling dangerously ill. This is not the case in the Roman period, however, when we can easily identify many amulets by their inscriptions and learn precisely who used them and why. A rather large chalcedony gemstone in the Getty Museum, for example, holds a short request: "Deliver Gaius from his fever, his chills, and all of his headache!" and a green jasper at the University of California at Berkeley is inscribed with another: "Turn aside the tension, the indigestion, the stomach-ache ofJulian, whom Nonna bore."2 These inscriptions inform us clearly about the names and genders of the patients and the nature of their illnesses, but little about their age. Although it is impossible to speak about pre-Roman amulets with the same degree of precision, we can use visual evidence to sketch a more general picture of the kinds of people who used amulets in the Classical and Hellenistic periods, especially as concerns their gender and age. In what follows, then, women and children will take first place in our inquiry, because, as was true for entry into the lifeboats of the

28

Archaeology

Distribution

sinking Titanic, these two social groups were apparently deemed by adult males to be the most fragile and thereby in the greatest need of protection and healing. And, indeed, classical vase paintings and votive statues, our earliest evidence for the deployment of Greek amulets, often depict women and children wearing amulets, but almost never adult men. These early amulets, moreover, often seem quite simple: a string with one or more attachments, much like a modern charm bracelet. We shall also see in this chapter how in the past, scholars have tended to marginalize this visual evidence for Greek amulets by dismissing them as the result of foreign (e.g., Phoenician) influence, as the reaction to some unique catastrophe (e.g., the Athenian plague), or as the practice of a specialized and morally suspect group (e.g., prostitutes). I will suggest that these attempts to isolate the use of amulets from mainstream classical Greek culture arise from the absence-outside of comedy-of literary evidence,3 as well as the discomfort some still feel about attributing magical practices or superstitious ideas to the Greeks of the classical period and most especially to the Athenians, a discomfort that the ever-skeptical Theophrastus himself seems to feel when he interprets the gesture of the dying Pericles as an admission of folly.4

1.1 Childhood Amulets in Classical Athens

Figure 1.1 Fragment of a fifth-century white-ground lekythos;British Museum, inv. GR 1905.7-10.10 used with the permission of the Trustees of the British Museum.

Naked male babies and young boys appear frequently on Athenian vase paintings of classical date with a string tied diagonally, like a bandolier, between one armpit

on miniature ritual jugs called choes (singular: chous) (Figure 1.2). These small

and the opposite shoulder, as we can see on a lovely fragment of a fifth-century white-ground lekythosthat was probably used as a funerary offering for the woman

jugs seem in one way or another to be mementos of an important religious feast called Choes ("Jugs"), which occurred on the final day of the Anthesteria, a three-

depicted on the right (Figure 1.1).5 The baby boy wears a string bracelet on his right wrist and a cord tied across his chest, from which loops and other small ob-

day wine festival dedicated to Dionysus. 9 This event, part ritual and part drinking

jects descend. An oft-cited study infers that the primary magical act behind these diagonal strings of amulets was the tying of knots that sympathetically bound hostile demons or diseases. 6 This seems to be a standard feature of Egyptian and Near Eastern amulets,7 but the early Greek evidence, as we shall see, generally does not support this inference, because the artists are often careful to show that small

party, seems to mark a major milestone in the life cycle for Athenian boys, 10 a time when they left the women's quarters and were handed over to their tutors. 11 At this festival, the male relatives of the extended family met and intoxicated themselves with wine drunk from their individual jugs. Young boys who attended the festival for the first time were apparently given miniature versions of the same

objects are attached to the string, which thus serves as a support or vehicle for

jugs (Figure 1.3), which often depict boys or male babies wearing the bandolier amulet string and in some cases thick bracelets on their right wrists that have been

other amulets, but is not an amulet itself. The diagonal amulet cord itself seems designed especially for children, and any parent can tell you why: it is nearly im-

added in white paint, suggesting they were of cloth or wool, similar to the "the bits of colored thread round wrists, arms, and necks" of children that the Chris-

possible for a child to shake off such a cord or choke on it, whereas an amulet attached solely to the neck would present a serious threat. 8

tian bishop Basilius complained about many centuries later. 12 Suspended from the chest we often see geometric forms similar to those on the funerary vase discussed

That such amulets were fairly typical for Athenian children and not simply the habit of a few superstitious families is suggested by their frequent appearance

earlier: crescents, disks, or circles and what appear to be short sticks tied in the middle. 13

29

30

Archaeology

Distribution

Figure 1.2 Attic miniature choesjug; Munich Antikensammlungen SH 2459 used with the permission of the museum.

One might at this point conclude that there existed an old Athenian tradition of baby amulets strung diagonally across the chest, but in recent years some scholars have suggested that the amulets depicted on the Anthesteria jugs, the major-

Figure 1.3 Birth ofErichthonius; British Museum, inv. GR 1837.6-9.54 used with the permission of the Trustees of the British Museum.

ity of which date to the last quarter of the fifth century BCE, were a unique and therefore untraditional reaction to the many deaths that occurred during the plague that killed Pericles and that they reflect a heightened anxiety about the survival of children in Athens. 14 It seems logical that anxiety about mortality did increase

cult statue of Hephaestus (Erichthonius' father), from which it was replicated on a

during the famous epidemic, but it is unlikely that the wearing of childhood am-

stage, Gaia emerges from the ground holding the infant king, who wears a string of amulets from armpit to shoulder, precisely like the ones worn by male children

ulets first became popular at this time, because even without the plague, the premature death of children was a great problem in antiquity, with rates running perhaps as high as 30 percent for infants and 40 percent for children under five years of age.15 The plague, moreover, cannot explain the same childhood amulet

series of Attic vases, beginning in the period of the Persian Wars. 18 At center

on the other vases. 19 It is not, I think, a coincidence that these miniature wine jugs, which mark a boy's first appearance in the ranks of Athenian citizens, depict boys wearing the

strings that appear throughout the fifth century on funerary lekythoiand in the scenes of the birth of their first king, Erichthonius (Plate 3),16 a story closely con-

same amulets as their first king and putative ancestor. Moreover, aside from his alleged role in founding the Panathenaic festival (in fact, a Peisistratid inven-

nected with Athenian claims-nearly unique among the Greeks-that they were an autochthonous people who had always lived in Attica.17 According to the traditional story, soon after Erichthonius was born, his mother Gaia ("Earth") handed

tion), 20 there are no important myths about Erichthonius-no military, heroic, or amorous exploits, in fact no adult experiences at all:21 he sprang from Athenian

him over to Athena, a scene that was sculpted in relief on the base of the Athenian

some sort and gave him to the daughters of Kekrops for safekeeping. The latter

soil wearing typically Athenian amulets, and then Athena placed him in a box of

31

32

Archaeology

Distribution

myth was recalled each year during the mysterious ritual of the Arrhephoria,

Etruria do, when they wear the Italic equivalent of the Greek amulet cord, namely

when a select group of Athenian maidens retraced the footsteps of Kekrops' hap-

the bulla,which was, however, worn on a necklace and not on a bandolier. 30

less daughters. 22 According to the story, Athena placed a pair of snakes in the

Similar votive statues with amulets like these have indeed been found through-

box to protect the baby from danger, creating a precedent for yet another traditional amulet reported by Euripides: snake necklaces of beaten gold (see Section

out the Greek world in the late classical period. In Argos, two small examples were discovered in a fifth-century tomb, 31 and earrings from South Russia depict boys

1.5). Erichthonius' birth, in other words, provides etiologies for one important civic ritual-the Arrhephoria-and for at least two different kinds of childhood amulets. 23

with amulets and exposed genitalia, suggesting that similar votives were dedicated in the Greek cities of the Black Sea.32 Early examples appear in western Greece as well, including a pair of miniature terracottas from Tauromina 33 and the stone torso of a boy from the sanctuary of Hera in Selinus, whose attachments include a horizontal tube on his chest and a disk or ring high on his shoulder. 34 Other mythical

1.2 Childhood Amulets in the Wider Greek World

heroes are also shown wearing these diagonal amulet cords. A gold coin of Tarentum (c. 340) depicts the city's eponymous hero Taras, like Erichthonius, as a child

The practice of attaching diagonal amulet cords to children is attested in the fifthcentury BCE in a number of other places in the Greek world that were unaffected

wearing a string of amulets, as he appeals to his father Poseidon, much the same as

by the plague. Similar amulet strings appear on limestone and terracotta votives

threatening snakes.35 More helpful, perhaps, because of the details it offers, is a standing marble statuette in the Vatican that appears to be a Roman copy of a Hel-

from western Greece, the Peloponnese, and especially Cyprus, such as we see on the chest of a small votive statue of a so-called temple boy now in the Metropolitan Museum of Art (Plate 3).24 This child wears a diagonal amulet cord like those of the Athenian boys, and here the three-dimensional medium allows us to see more clearly the attachments as rings, disks, and crescent moons, as well as verti-

the baby Heracles on a contemporary Cypriot plaque wears one as he strangles two

lenistic original (Figure 1.4).36 This boy has amulets suspended at every point along the string on both his chest and back, amulets that we can see more clearly on the numbered drawing in Figure 1.5: trefoils (nos. 2, 6, 15, and 18), crescent moons (3,

cally and horizontally suspended rectangles or cylinders. Note, too, that the boy

10, and 16), double axes (4, 14, and 17), leaves or spearheads (5, 9, 13, and 20), an open hand (8), or a dolphin (11).37A good number of similarly outfitted votives were

wears thick bracelets on the left arm, again like some of the Athenian children.

discovered in Paestum in an Italic shrine of Mater Matuta, a cache that probably

The Cypriot boys differ from the Athenian ones in one important way, however: they are usually depicted wearing an amulet cord over their clothes, although in

dates to the third century BCE (Figure 1.6).38 Like the Cypriot boys, they wear

many cases the bottom edge of the tunic is lifted to expose their genitals. 25 Many of these statuettes were discovered in sanctuaries of Apollo, Aphrodite, and other kourotrophic divinities on Cyprus, and thus they seem to have served as votives,

their chest amulets outside of their clothing, although these votives apparently depict much younger children-infants, in fact-in swaddling clothes. These swaddled images are a widespread phenomenon on the Italian peninsula, and many wear the traditionally Italic bulla on a necklace over their bands. Only the terracottas

presumably thanking the god of the sanctuary and celebrating the survival of a male child during a crucial transition in his life, when the boy would move from the women's quarters to the men's26 -that is, at the same time suggested by the

from Paestum, however, display diagonal amulet cords, leading to the correct infer-

Athenian choes.But because of the strong Phoenician presence on Cyprus prior to

see the same variety of attachments to the main cord, including the crescent moon, the disk, and what appears to be a double axe without a handle. 40

the arrival of the Greeks, some scholars have suspected that these amulets reflect an eastern, non-Greek practice, and as a result, few have emphasized, at least until recently, the obvious parallels between the Cypriot amulets and those on the Athenian vases.27 A Phoenician origin for the amulet cords is, however, a mirage. Al-

ence that in this detail, at least, they reflect the persistence of an old Greek tradition in a city that was in fact founded by the Greeks. 39 On the Paestum votives, too, we

Taken together, the visual evidence for such amulet cords from the Peloponnese, the Black Sea, Cyprus, and western Greece, along with the Vatican statue and the vase paintings from Attica, confirms the suggestion that diagonal strings

though it is true that the crouching posture itself of many of these boys seems to have been invented in Egypt and transmitted to the west by the Phoenicians, 28 it

of amulets were worn by male babies and young boys in various places in the Greekspeaking world throughout the classical and Hellenistic periods. 41 The general

is only in the Greek world that these boys wear diagonal strings of amulets. 29 These cords, in short, reflect a local Greek practice, just as the crouching figures from

absence, moreover, of amulet strings of the diagonal type on native Egyptian, Near Eastern, and Italic statues and the high degree of Greek influence in all of the

33

34

Archaeology

Distribution

J

~ ·~

.; J

.:>J~~.·' ·--.. ,....,

Figure 1.5 Amulet string across the chest and back of a Roman-period marble copy of a Hellenistic statuette in the Vatican; after Diilger (1932).

parents. The exaggerated display of the male genitalia on many of these votives as well as the presence of amulet cords on "founding fathers" like Erichthonius or Taras suggest another important goal: the forward march to manhood and

)1i

citizenship.

'illf Figure 1.4 Roman-period marble copy of a Hellenistic statuette in the Vatican; after Visconti (1849) fig. 22.

1.3 Amulets for Courtesans and Bathers Naked women also reveal their amulets in Attic vase paintings of classical date, albeit in far fewer numbers. When these women show up on wine cups and pitchers,

areas where they do appear suggest that even in places of strong Phoenician or Italic influence, such as Cyprus or Paestum, this particular vehicle for amulets re-

scholars are undoubtedly correct to call them flute girls or prostitutes, especially

flects a Greek practice. It is also of great interest that so many of these statues

when they drink wine or play music, as they do in the scene in Figure 1.7A, where the woman on the right has an object tied to her right thigh and the one on the left

have been found in temples, either of gods like Apollo, who in Cyprus was the patron deity of boys and young men, or of kourotrophic divinities like Hera on

has a similar device on her right forearm. A flute girl from the same vase (Figure 1.7B)has a similar device on her upper right arm. Since none of these attachments

the West Hill at Selinus or Mater Matuta in Paestum. 42 We saw that at Athens, too, the Anthesteria jugs performed a similarly commemorative function under the

corresponds directly to any traditional form of Greek jewelry,43 scholars usually

aegis of Dionysus. It seems clear that these images in one way or another marked the end of the dangerous period of infancy and provide a snapshot, as it were, of a healthy male baby or young boy who has been successfully protected by his

assume they are amulets, and indeed they have dubbed the one strapped to the thigh a "garter amulet." 44 In some bathing or dressing scenes, however, women also wear amulets-for example, on the red-figured vase in Munich in Figure 1.8, where we see the woman on the right wearing a garter amulet, while the one in

35

36

Distribution

Archaeology

Figure 1.7 A and B Courtesans on an Attic red-figure cup; after Jahn (1867) fig. 1. Figure 1.6 Votive statuette from the shrine of Mater Matuta in Paestum; photo courtesy of the photographic department of the Archeological Park of Paestum; photographer: Francesco Valletta.

vases are courtesans. First, since Aphrodite alone of the major goddesses is depicted nude, we have no idea what Athena or Hera wear under their clothes. Second, women sometimes wear garter amulets in bathing or dressing scenes where there

the center has around her neck a single pendant on a simple string, which has also been called an amulet.

45

Unlike the hundreds of images of boys wearing amulets,

are no wine cups or sexually aroused men, so there is nothing to suggest that these women are anyone other than proper wives or daughters.

50

The crucial de-

there are only a dozen or so of grown women, with roughly half of them depicting

terminant is the type of vase on which the image appears: if it is a wine cup or any

courtesans drinking or dancing and half of them in domestic scenes dressing or

other ceramic ware used at the symposium, it is highly likely that the women

bathing.

46

The more detailed examples show rings, crescents, or a small club as the 47

bathing scene associated with a woman's ceremonial bath before her wedding, these

single attachment (Figure 1.9).

Until recently, scholars generally assumed that all women wearing garter amulets in Greek vase paintings were prostitutes and that these amulets must therefore be contraceptive or otherwise associated with their lifestyle as sex workers, an inference that seems supported by the fact that Aphrodite and Eros alone of the Greek gods are occasionally shown wearing the same kind of thigh amulet.

depicted are indeed some kind of sex worker, but when nude women appear in a

48

images could just as well be the idealized views of a young citizen daughter and her friends on the eve of her wedding. 51 Later literary evidence confirms that in the ancient world women used waist and thigh amulets for specifically female situations, especially to ease childbirth, to regulate menstruation, or to prevent spontaneous abortion.

52

The evidence for

There is also a well-excavated house just inside of the Sacred Gate at Athens that

birthing amulets is even more extensive. A Hippocratic text directs us to smear

seems to have been inhabited in the fourth century BCE by hetaerae or flute girls

vegetal matter on wax, wind it up in a tuft of red-colored fleece, and tie it around

who did indeed wear disk amulets, including one of Aphrodite riding a goat (see

the loin of a female patient.

next section).

49

But it is not entirely clear that all of these nude women on Attic

53

Several centuries later, Xenocrates, apparently draw-

ing on a Mesopotamian tradition, discusses the similar use of the "eagle stone" on

37

38

Archaeology

the left thigh,

54

while Dioscorides in his work on medical drugs mentions that

"everyone thinks that jasper stones ... , when tied around the thigh, were promoters of quick birth (okutokia)."55 The lapidaries have similar instructions placement-for

for

example, placing the "eagle stone" on the kidneys of a woman or

tying a stone called exebenusto her right thigh in order to promote a quick or easy birth. Their instructions for the use of galactite to promote lactation are more detailed: attach the stone to the left thigh with a wool thread from a fecund sheep. 56 Roman-period and later recipes for amulets to control bleeding often distinguish the gender of the patient by the placement of the amulet. A Latin recipe to stop bleeding tells us, for example, to write a special text on a sheet of papyrus and then tie it with a thread to the sufferer's body-around

the neck for men and the

belly for women. 57 Only rarely are thigh amulets designed for both sexes, mainly for diseases of the lower abdomen, such as hemorrhoids or colic. 58 The short stick tied in the middle of the upper arm of the courtesan illustrated in Figure 1.7B bears witness, however, to an even wider tradition of curative amulets for both sexes. A papyrus amulet for a woman named Arete suffering from fever mistakenly preserves at its end the handbook instructions: "Tie the strip to the right arm with a holy thread from a garment, and it ceases altogether." 59 And Dioscorides mentions different faunal substances, which, if wrapped in pouches and tied to the arm, will cure rabies, epilepsy, and quartan fever.60 He does not specify the right or left arm, but when the lapidaries recommend arm amulets, they are usually tied to the left. An agate engraved with the Roman god Mercury, for instance, when placed on the left arm or neck, brings good business; the red stone from the entrails of a swallow wrapped in linen and tied to the left arm cures lunacy and-if

the linen is dyed with saffron-fevers

and tumors; and another

Figure 1.8 Female bathers wearing amulets on an Attic red-figured stamnos;Munich Antikensammlungen; after Ohly-Dumm und Hamdorf (1981) 66 Tafel 28 used with the permission of the museum.

stone tied to the left arm protects against powerful people and all trouble. 61 And some arm amulets were designed explicitly for women: the "eagle stone" (mentioned earlier), if tied around the left arm, was thought to prevent false labor, while the dock-plant tied on the left wrist cured childlessness. 62 Among the recipes collected by Pliny for medical amulets, more are placed on the arm than on the neck or thigh-even

those for gynecological problems. 63 But the recipes need not

be medical: a recently published papyrus recipe for a charm to restrain anger tells us, "Take a slip of papyrus and write on it kei pherpheret... and tie it around your left arm." 64 These practices continued well into Christian times. In the mid-eighth century CE, Boniface complains to the Pope that women in Rome "were in pagan fashion wearing amulets tied to their arms and legs and even offering them openly for sale."65 It is unclear, however, how many women wore arm or garter amulets or how often they did so. In the vase painting of the courtesans (Figure 1.7Aand B), only three of the four women are wearing them, and on the Munich vase (Figure 1.8)

Figure 1.9 Female bather wearing garter amulet on an Attic red-figured squat lelqthos(430-420 BCE); Metropolitan Museum of Art, acc. no. 30.11.8; after MMA Bulletin 25 (1930) fig. 2.

40

Distribution

Archaeology

only two of three. These limited appearances suggest they either were not worn by all women or were worn only intermittently.

1.4 Circles and Crescent Moons Geometric shapes, especially circles and crescent moons, are depicted most often on the amulet cords in classical-period images of women and boys. The circular amulet seems to have had a long history in the ancient Near Eastern and Mediterranean worlds-for

example, solar disks, wheels, bullae, or repurposed coins. 66

Undoubtedly the most popular in the early Greek world was the circular type of gorgoneion(Figure 1.10). The goddess Athena from earliest times terrifies her personal enemies and those of her human proteges with a circular and medallion-like gorgoneionplaced in the middle of her chest or her shield. 67 This image of the disembodied head with snaky hair appears frequently in archaic Greek art on city walls, armor, shields, and chariots, most of the time placed in some obviously defensive or apotropaic position. 68 In the same period, gorgoneiaalso appear on metal jewelry and terracotta vessels (including small perfume containers) and on public buildings in the form of acroteria and antefixes. 69 Although a handful of early Gorgon heads are depicted with a body-for example, the figure on the famous temple pediment on Corfu-such scenes are statistically quite rare, and most would nowadays agree that the story of Medusa's decapitation was created to ex70 plain the existence and protective power of the gorgoneion. Indeed, Homer, our earliest literary source, never refers to Medusa by name or to her story, but only to the Gorgon's head, a thing of horror with distinctive eyes71 that always appears in contexts where its apotropaic power72 is most obvious, as in the center of a military breastplate or shield.73 The gorgoneiaof archaic and early classical periods terrify because they are a hybrid mix of different wild animals, usually of a predatory nature, in which the lion predominates. 74 Circular gorgoneia remain popular throughout antiquity. In Etruria, they decorate bullaeand clearly add to the protective power of this Italic amulet container-for example, on a fourth-century BCE piece from Chiusi that also has an animal's eyetooth hanging from it. 75 And in a residential section of late fourth-century BCE Gela, archaeologists unearthed hundreds of small terracotta disks (oscilla)with suspension holes decorated primarily with satyr heads, gorgoneia,and pretty female faces with peculiar melon hairstyles that scholars associate with sirens. 76 In later times these images became prettier and were hung at the tops of entranceways or porticoes, but in fourthcentury Sicily they also seem to have been used as loom weights and thus they probably protected the weaving of women. 77 In the Roman period, gorgoneiaare ubiquitous on small gold disks, on larger medallions, engraved in red-jasper

Figure 1.10 Black-figured gorgoneion(Cabinet des Medailles); afi:er de Ridder (1902) p. 220 no. 232.

intaglios (Plate 4), or sculpted in high relief in jet and bone (Figure 1.11).78 One lapidary recipe boasts that when this image is carved into red coral, it is the "greatest phylactery," and indeed examples survive in both red coral and in red jasper (see Figures 3.5A and B).79 The parallel appearances of these frontal faces on temple antefixes, domestic oscilla,personal jewelry, and military armor, suggest, moreover, an obvious assimilation between buildings and bodies and their need to be protected. Gods are also depicted on circular medallions that seem to have amuletic power. Aphrodite "upon the goat" (Epitragia), for example, appears on a series of metal disks found in late classical houses and graves that show her seated "side-saddle" on a goat with stars or a crescent moon in the background. The earliest and largest is a silver example found with other amulets in the late fourth-century BCE "Building Z" located just inside the Sacred Gate at Athens (Figure 1.12).80 In the background, in addition to the moon and stars, we see two running kids (in the caprian sense) and two naked young males, Eros to the right crowning his mother and on the left a torchbearer, who may be Phosphorous, leading the way. As we shall see later, the presence of the moon and stars on a gemstone usually triggers the suspicion

41

42

Archaeology

Distribution

Figure 1.11 Gorgoneionpendant carved in onyx from Rome; after Goodenough (1953)no. 1044.

that it is an amulet, and this is probably the case here, although only the addition of text would prove this definitively. 81 The image of Aphrodite and the goat seems to be the result of a fourth-century

synthesis of the Athenian iconography of

Aphrodite Pandemos with that of the goddess as the evening star, one that was perhaps also present in a famous, but no longer extant, fourth-century statue of the goddess carved by Scopas. 82 At Athens, Aphrodite Pandemos was the patroness of courtesans and, because the layout of "Building Z" recalls a tavern, excavators have plausibly suggested that the owner of the amulet may have indeed been a courtesan.

83

At any rate, the owner belonged to a small group of women who

worshipped Cybele and wore a variety of other amulets, including a gorgoneion,

Figure 1.12 Silver medallion from Kerameikos depicting Aphrodite Epitragia; after Knigge (1982) fig. 86.

eye-beads of the Punic type, and a much smal !er silver medallion depicting a female head in profile with two stars and a crescent moon. Other circular images

a small lion standing before her right leg. It was found with an amulet string in a

of Epitragia, stamped in bronze, silver, or gold, have been found on Naxos and

mid-fifth-century BCE child's grave in Apollonia Pontica. 88

Delos and in Thessaly and Macedonia. All date to the Hellenistic period and often appear in domestic or funereal contexts, which include other amulets, especially lunulae.

84

Coins-especially

reused from a bygone era 89-were

another circular shape

used frequently as amulets. Our earliest evidence also comes from the late classi-

The Hellenistic grave of a wealthy woman in Thessaly yielded a

cal graves of children in Apollonia, where three pierced coins of silver or bronze

gold amulet cord with two lunulae and two medallions (Helios and Selene), as

were found as parts of different amulet strings that also included beads, pierced

well as circular images of Aphrodite Epitragia with stars or a crescent moon in the

shells, and pierced animal teeth. 90 Beginning in the Hellenistic period, we first

background.

85

As was mentioned in the introduction, in the later periods Helios

find evidence of coins mounted in gold or silver and worn as jewelry. 91 The use of

and Selene become important figures on amulets, and we can probably identify as

coins makes sense because they were already the right size and circular shape, and

an amulet at least one other medallion depicting the two of them: a gold medal-

beginning with the coins of Alexander the Great they usually displayed on the

lion of third-century CE date from eastern Greece that was later surmounted with

obverse a powerful man, often in armor and of imperial demeanor, if not divine. 92

a thin lunula and hung on a chain with two other items that may have had amu-

On the reverse, these coins often carry a scene associated with victory, first the

letic power, a acorn and a round-bellied pendant, both of gold. 86 Helios and Selene

goddess Nike herself ("Victory"), and later other images of imperial power, such

87

More difficult to explain is a

as Helios in his chariot or a mounted horseman spearing a supine figure. 93 We can

unique gilded silver medallion that shows Peleus carrying off Thetis, who has

see from the critics, too, that the image depicted on the coin was important. The

also appear together on a few gemstone pendants.

43

44

Distribution

Archaeology

Talmud notes that both the metal media and the image were worrisome, but allows gout sufferers to bind a coin to their leg on the Sabbath, while Chrysostom criticizes Christians in fourth-century Antioch who tie bronze coins of Alexander to their head and feet. He was annoyed that they were placing their hopes in an image of a king of the Hellenes rather than in one of Christ. 94 Coins were pierced in their upper rim or encased in a setting of gold or silver, so we know they were worn as ornaments, but how can we tell that they were amulets? Some have suggested that men or women who wore coins in late antiquity were motivated by the lack of carved gemstones, by the desire to flatter the emperor and his family, or by the need to hoard gold, when true coinage had become scarce. 95 In some cases, the addition of text or the alteration of a coin's iconography reveals its use as an amulet, 96 but most often the archaeological context reveals its purpose-for example, the coin of Trajan found inside a bulla in Hungary or an early Byzantine military helmet stamped with dies imitating coins and inscriptions asking for help and health. 97 Most telling are the coins that show up frequently in infant burials, where they are accompanied by other traditional amulets, for example, those from Apollonia described above, but also a pierced coin of Nero found with amber beads, a bronze bell, and a bone disk engraved with a phallus all hung together on a necklace from a Roman tomb that contained the skeletons of many infants. 98 The crescent moon also appears frequently on the amulet cords worn by women and children in the images discussed earlier, and, indeed, it is probably the most

Figure 1.13 Two child's lunulae;British Museum, Marshall (1911)nos. 2909 and 2911; author's photograph used with the permission of the Trustees of the British Museum.

widespread and long-lived geometric shape used as an amulet in the ancient world (Figure 1.13). For the Greeks, Hesychius' lexicon is quite explicit: "Crescent moon

(selenis)-a protective amulet (phulakterion)on the neck of children." It was also called a meniskosin Greek and lunula in Latin. We saw earlier, too, that Basilius,

included a bone amulet carved with a lunula.109 A question asked of a long-lost

a bishop of Caesarea, described the "moon-shaped plates of gold, silver or cheaper

young girls on their birthdays_11°Crescents were, moreover, sometimes attached to

material, which foolish old women fasten upon infants."

99

Iconography and archae-

ology confirm that lunulaewere worn much earlier in ancient Mesopotamia, Syria, and Palestine, but rarely in pre-Roman Egypt. 101

100

They begin to appear in the Greek

daughter in an early Roman comedy suggests, in fact, that fathers gave lunulae to the harnesses of valuable animals-for

example, the horses sculpted on Trajan's col-

umn.111 The visual evidence, temple inventories, and grave archaeology for the classical, Hellenistic, and Roman periods all suggest a similar pattern: women,

worn mainly by women, children, and prized

children, and large domestic animals were likely to wear lunulae, but adult men

animals. 102 Numerous silver, gold, and lead examples were dedicated in sanctuar-

were not. 112 Certainly the crescent moon makes good sense as an amulet concerned

ies of Demeter, Hera, or Artemis in the archaic and early classical periods 103 and

with a woman's fertility, the menstrual cycle, or perhaps even lunar worship, 113

others were buried with Greek women and children in the classical period at Er-

but the much earlier use of the lunula in Mesopotamia on an amulet cord for ep-

etria, Olynthus, and elsewhere. 104 They appear "everywhere in the Hellenistic world

ilepsy, a disease that throughout antiquity was thought to be connected with the

world by at least the Bronze Age,

as part of women's apparel"

105

and are equally ubiquitous in Roman times,

106 show-

ing up on women in the Palmyra funerary reliefs or in the Fayyum portraits (for

phases of the moon, suggests that there could be other reasons why children and animals wore them as well. 114

the latter, see Figure 1.16).107 They also appear in the Roman-period graves of both

The crescent moon, finally, often appears on other amulets with a star or a

women and children: a silver crescent was found in the grave of a child, and

group of stars. We saw them earlier, for example, on the Epitragia medallions (Fig-

the skeleton of an adult woman was unearthed wearing a string of glass beads with a silver lunula as the central ornament.

108

The cord of another child in N:i'mes

ure 1.12), and they are sometimes added to magical gems-for side of a divinity-or

example, on either

even appear alone. 115 We see them also on a small marble

45

46

Archaeology

Distribution

Roman times, from which period we are fortunate to have a handful of wellexcavated graveyards, as well as the Fayyum "mummy portraits," some of which wear on their necks single amulets, such as the gorgoneion(Figure 1.15), the lunula (Figure 1.16), or the horizontal amulet case (Figure 1.17), which will be discussed in the next chapter. These portraits give us a rough assessment of Greek practices in Egypt in the first three centuries CE. 122 Of the extant portraits, 134 wear on their necks objects that are generally recognized as amulets, and none of these are adult males, despite the fact that men make up a sizable proportion of the corpus. 123

Figure 1.14 Marble bust of Selene from Argos with hidden magical inscription; drawing after Delatte (1913a).

bas-relief from Argos (Figure 1.14) that depicts in an arched niche a frontal bust

Crescent moons (lunulae):

36 women and 1 boy

Horizontal amulet capsules:

3 women, 18 boys, and 1 girl

Bullae: Coins or medallions: Gorgoneia: Paired busts of Isis and Sarapis:

3 boys 32 women 17 women and 1 boy 5 boys

These figures undercount the total number of amulets, of course, because the portraits usually include only the head or bust of a fully dressed person and thus do not reveal any amulets worn beneath the clothes or on the lower body. But even in the limited case of neck amulets, they can tell us much about the relative distribution of amulets according to the gender and age of the wearer, information

of Selene with large eyes, whose pupils are strongly emphasized. 116 Three stars have

that coincides, for instance, with our expectations that the crescent moons were

been added on either side of her head and on top a crescent moon with horns pointed up-Selene's usual crown-within which sits the seventh star. Although

worn primarily by female adults and the bulla by male children. The restriction of

scholars originally thought the image was a votive or the focus of a cult in a communal hall, 117 the magical nonsense inscription engraved on the underside of the bust and thus completely hidden from view118 suggests instead that the image was designed to protect the building or room in which it stood, "empowered by the inscription concealed on the bottom." 119 It works, in short, much the same as a gem amulet that depicts a divinity on one side and on the other a magical name or symbol hidden from the casual viewer (see, e.g., Figure 0.4). Selene only rarely appears on the extant magical gems, 120 but a lapidary recipe directs us to inscribe a "moonstone" with a bust of Selene that will serve as a charisma amulet. 121

1.5 Comparative Data from the Roman Period The age and gender of those who wear amulets in the visual evidence of the classical and Hellenistic periods correspond well with the distribution of amulets in

gorgoneiaand coins or medallions to women is, however, somewhat surprising, since the former seems to have been a universal apotropaionand examples of the latter, pierced for suspension, do show up in the tombs of children. 124 The high proportion of boys wearing horizontal amulet cases suggests, moreover, that contrary to scholarly assumptions, some, if not many of the gold and silver foil amulets (lamellae) inscribed with the names of male patients may have been commissioned for male children rather than adult men. 125 Nearly all of the amulets on the mummy portraits seem, finally, to be suspended singly on a metal chain or leather cord, even those that appear on children, although this does not rule out the possibility that the children also wore a fully loaded amulet cord underneath their tunic. Recent work on Roman-period cemeteries confirms this pattern of primary concern for women and children. A study of small glass beads found in Roman imperial graves, where the age and sex of the deceased were known, concluded, among other things, that roughly 40 percent were found in the graves of adult females and the rest in children's graves. In cases where amber beads were included, the ratio of difference was even greater: 20 percent adult females to 80 percent

47

48

Distribution

Archaeology

Figure 1.15 Fayyum portrait of woman with gorgoneionmedallion; after Hilton Price (1908) no. 4748; drawing by E. L. H. Cooper.

children. 126 There were, however, no such beads in graves where the deceased could be identified as an adult male. One could argue of course, that the beads found in female graves were solely for adornment, but one cannot so easily make the same argument for small children of both genders. The presence of the amber beads is especially important because amber was widely recognized as having curative and

Figure 1.16 Fayyum portrait of woman with lunula; after Parlasca (1969)no. 247.

protective powers, especially for women and children. Pliny, speaking of the Po Valley, says that amber was worn by both: "Even today the peasant women ... wear pieces of amber as necklaces, chiefly as an ornament, but also as a remedy because

of a man or an animal and amber on the neck of a woman. 128 A survey of small

amber is, indeed, believed to be a prophylactic good against tonsillitis and other

bronze bullaein the Roman cemeteries of Hungary came to similar conclusions

affiictions of the throat"; he goes on to add that "amber ... is a benefit to babies,

about gender and age. Of the twenty-five graves containing bullae,where the age

when it is attached to them as an amulet."

127

St. Eloi confirms the special focus on

and sex of the skeleton could be assessed, there were five adult women, fifteen girls,

women when he forbids early Christians in Gaul to hang phylacteries on the neck

three boys, but only two adult males. 129 This study is especially interesting given

49

so

Distribution

Archaeology

to her son when she exposed him as a child: "Serpents, Athena's gift all gold; she bids us raise our children in these in imitation of ancient Erichthonius ....

Neck-

laces for the new-born baby to wear."131 These lines in the recognition scene at the play's finale are important, because they echo those of Hermes in the prologue: "For Zeus' daughter gave to him (i.e. Erichthonius) two serpents to guard his body....

And that is why the Athenians have the custom of rearing their children

in snakes of beaten gold." 132 It is curious, however, that in the first passage Euripides speaks of neck ornaments, whereas in the archaeological evidence, snakes almost always appear on bracelets and rings, rather than necklaces. Indeed, the snake has been described as the most popular form of bracelet in ancient Greece, and it appears in two common types: a Hellenistic and early Roman type that depict a single snake with a tail, sometimes coiled more than once around the wrist, and a slightly later and simpler type of open bracelet with a pair of facing snake-head finials. The latter clearly depicts the head of a snake of the colubridaspecies connected with the worship of Asclepius. 133 Scholars have long suspected that the latter was worn as a protective amulet, 134 but no text tells us this explicitly until late antiquity, when a bronze bracelet discovered at Olympia was inscribed with a protective inscription written in two parts, each close to one of the snake heads: "O Lord, rescue her who wears [it]."135 In the graves of Olynthus and Apollonia Pontica, simple, undecorated bracelets were found most often in the graves of small children, some of which were used as anklets; one of the bronze anklets had opposing snake heads. 136 We find finger rings in both styles beginning in the Hellenistic period, and a recent study of the rings found in the jewelry hoards of Britain shows that by the Roman period, the second type-the snake-heads-had

open ring with opposed

become the most popular and continued to be so down to the

end of the empire. 137 Figure 1.17 Fayyum portrait of boy with amulet case; after Parlasca (1969) no. 35.

1.6 Summary There is abundant visual evidence for Greek amulets prior to the arrival of the Romans in the eastern Mediterranean, and their use is clearly dictated by the gen-

the fact that Roman literary sources frequently mention the bullaas a sign of male

der and age of the owners. Amulets attached to adult female bodies differ from

status or even citizenship, but only rarely mention females wearing them. Here, as

childhood amulets in their number and placement: women seem to wear only one

is so often the case, it seems that literary texts focus exclusively on how, when,

or two at a time, whereas children could sometimes be loaded down with a dozen

and why elite boys wear the bulla, and almost entirely ignore the fact that girls

or more. This suggests that childhood amulets, much like the battery of inocula-

and grown women could wear them as well.

130

tions given to babies in modern times, were used comprehensively and prophylac-

Scholars have noted the extraordinary popularity throughout antiquity of snake

tically against many potential dangers. Women, on the other hand, wore fewer

jewelry on women and children, and some have understandably connected this tra-

amulets and did so sporadically, presumably for curative and other reasons. To my

dition to Creusa's explanation in Euripides' Ion for the gold snakes that she gave

knowledge, no pre-Roman Greek text makes mention of amulets worn by children,

51

52

Archaeology

with the exception of the recognition scene in Euripides' Ion. 138 The visual evidence for childhood amulets is primarily votive or commemorative. The Anthes-

Distribution

both men and women to cure health problems or bring some abstract benefit. There

teria jugs in Athens and the votive statuettes from the Peloponnese, Cyprus, and

is, however, almost no iconographic evidence that adult males wore amulets in the ancient Greek world, a fact that cannot be so easily explained by the uneven sur-

western Greece all presumably celebrate a successful passage through the dangers of infancy-a success helped, these images seem to say, by the many amulets de-

vival of evidence. Indeed, given the ideal of heroic nudity in the Greek depiction of the adult male body,143we would suspect that if grown males did wear amulets

picted. The parallel with the Roman bulla is again instructive, because literary

in classical period, they would have shown up on the innumerable male nudes in vase paintings or sculpture. 144The anecdote about the last days of Pericles quoted

sources tell us that bulla was indeed put off when the boy was recognized as a man. 139If we assume that the child put off his amulets at the age he is depicted in

at the start of the chapter tells us, however, that adult males did indeed don am-

the votive, the younger age of the children in the Greek evidence suggests that

ulets if they were ill, and indeed, the widespread evidence for medical amulets sur-

this perceived period of danger or vulnerability may have been of shorter duration in Greek cities than in Rome and even shorter still at Paestum, where we find swaddled infants with similar amulet cords. 140

veyed in the previous chapter suggests that sick or injured people, regardless of

The depiction of the infant king Erichthonius and the eponymous hero of Tarentum with similar kinds of diagonal amulet cords suggest that Greek child-

that athletes wear amulets, but these seem to have been of a different type aimed

hood amulets were also important markers of citizenship or social class. Indeed,

in the know recommend that we carry the gizzard stones of roosters in our mouth

although more than half of the surviving Anthesteria vases show boys wearing amulet strings, that still leaves a good number of have-nots. There is a similar

or in a leather pouch on the arm in order to fight without fatigue or thirst, adding that to prevent the loss of the amulet, it should be placed in a durable covering, but one that does not raise suspicion. 145He says that it can also be used by sol-

ratio among the Cypriot temple boys: a majority of them wear amulets, but many do not. Here, too, it is impossible to separate completely two important features of these amulets: they seem to mark a male child as one especially protected by his parents, but also as one who has parents of an elite political or economic status. In Athens and Tarentum, the amulet cord may even have acted as a sign of citizenship, by recalling the childhood amulets of their ancestor or founding father. Some have suggested, too, that the signet rings worn on the amulet cords of some Cypriot temple boys, in addition to serving as amulets, also marked these children as descendants of powerful officials who once owned the individual rings. 141Here,

gender, wore curative amulets. But what about protective amulets? A scant handful oflater sources do tell us at procuring victory rather than protection. Africanus, for example, says that those

diers, athletes, and gladiators, but his comment about raising suspicion suggests he is mainly concerned with the latter two, an idea borne out by an anecdote preserved by Photius, where he discusses a famous protective incantation known as the Ephesiagrammata ("Ephesian letters"): "When a Milesian and an Ephesian were wrestling at Olympia, the Milesian was unable to wrestle successfully, because the other man had the Ephesiagrammata around his ankle. When this became known and they were removed, the Ephesian was defeated thirty times in succession."146 Africanus' fear of suspicion and the outcome of Photius' tale both suggest, of course,

again, the Roman bullaprovides an interesting parallel, because our literary sources suggest that this ornament was also a mark of distinction: elite children wore a

that there was then, as in the present day with regard to performance-enhancing drugs, a general approbation against using amulets in such competitions and that

gold bulla as an amulet container, while others of lower birth wore a leather case

ancient Greek athletes, like modern ones, sometimes ignored it in their desire to win. There is also the evidence of the victory amulets, 147such as the papyrus that asks "Give victory and safety at the stadium and in the crowd to ... Sarapammon!" or the advice in a handbook to inscribe this short prayer on a runner's big toenail: "Grant success ... and victory in the stadium!" 148In the latter case, like that of

called the lorum. 142 The powerful substance or object sealed within both could presumably be the same; the type of container, however, marked the owner's wealth and social status. The visual evidence for amulets on grown women is less abundant than for boys, most probably because of the widespread Greek scruple in classical times against depicting nude females. But the vase paintings are informative. The sex workers and bathing parties seem to wear similar kinds of amulets, including the "garter amulet" and another kind tied to the arm. The intermittent appearance of

the Ephesian amulet discussed by Photius, the amulet is hidden from view. Other examples of secret amulets include one placed in a sandal and another under a garment.149Battle amulets, on the other hand, are surprisingly difficult to locate in our literary evidence,150 despite the obvious dangers of warfare. 151It would seem,

these amulets, however, makes it unlikely that all women wore amulets all the time,

then, that men did wear amulets to cure disease and encourage victory, but only

and later handbooks suggest that thigh amulets were used primarily to regulate menstrual cycles and to help in birthing, whereas arm amulets could be used by

rarely to protect themselves.

53

Shapes

CHAPTER 2

Shapes

The vase paintings of the classical era gave us the general impression that the cords worn by women and children were strung with individual amulets that could be distinguished by their shapes, the two most common being the circle and the crescent.1 And as we saw with regard to crescents and circles, our best evidence for these individual shapes are the remains of the amulets themselves recovered from late classical and Hellenistic graves, some carved in amber or bone and others cast from valuable metals or formed by pressing metal foil into a mold. Although the diagonal cord itself was, as we saw, a peculiarly Greek practice, a number of these individual shapes appear much earlier in the Near East. In Mesopotamia, for ex-

Figure 2.1 Necklace from Crimea; after Reinach (1892) 51 no. 3, plate 9.3.

ample, men and women alike wore amulets, usually a beaded string on the neck, wrist, or ankle that could also include disks, lunulae, shells, and beads. 2 Middle

principle in the identification of amulets in the graves of children or on the neck-

Kingdom recipes from Pharaonic Egypt seem to reflect a similar tradition, such as

laces of grown women is their eclectic nature: unlike ornamental necklaces that

an incantation to be sung over beads of gold and carnelian, a seal, a crocodile, and

usually include a limited number of repeated shapes arranged in a pleasing man-

an open hand, which are all strung on a thread of fine linen and applied to the

ner, amulet strings, like modern charm bracelets, seem to incorporate many dif-

throat of a sick child. 3 One could also string a faience lion on a red cord and at-

ferent shapes, that are usually not repeated and not arranged in any discernable

tach it to the hand to protect against snakes in the bedroom, presumably while

pattern.

4

sleeping. From a pair of Phoenician tombs of the seventh and sixth centuries BCE, we find similar assemblages: in the first were found twenty cowrie shells on a string with a swan-headed silver amulet case, a blue enamel scarab, a silver disk, a blue

2.1 Amulets Found in Graves

eye-bead, and two bronze bells; and in the second, more opulent tomb were a heart of carnelian, a gold and rock-crystal acorn, a gold amulet tube, a scarab and other shapes. 5

A few well-preserved assemblages from late classical Greek graves reveal individ-

In what follows, we will continue to examine the archaeological evidence for

century grave at Abdera include a hand in the obscene gesture of the flea (discussed

ual amulets of great variety. Twenty-one bone beads discovered in the fourth-

the individual shapes used as three-dimensional amulets in Greece before the Ro-

in Section 2.3), a scallop seashell, a double axe, a trefoil, a dolphin, and a cicada,

man period, moving away from the geometric shapes surveyed in the preceding

and at Acanthus a tomb of similar date for a three-year-old child yielded a string

chapter to discuss those that take the form of weapons, seashells, and animals, as

of small gold-foil amulets produced from a mold: a club of Heracles, a sea shell, a

well as those that take the shape of human body parts: the frontal face, the eye,

double axe, an open hand (vertical), the frontal face of a satyr, and a tortoise. 6 The

the hand in various gestures, and the private parts, both male and female. A guiding

remains of a late classical or early Hellenistic funeral pyre on the island of Thasos

55

56

Shapes

Archaeology

57

:--r=-.-,"" Figure 2.3 Line drawing of a painting on a red-figure pyxis from a fourth-century BCE Attic workshop; Benaki Museum; after Ziva (2009) fig. 25 with thanks to the museum for providing the original artwork.

including from left to right a trefoil, a double axe, an open hand (horizontal), a dolphin, a lunula, a second trefoil, a cicada, a ship's ram, a frontal eye, a grasshopper, and a peapod. Since a pyxis often served as a jewelry box, this painted amulet Figure 2.2 Necklace from Crimea; after Reinach (1892) 51 no. 1, plate 11.1.

cord represents either the daily jewelry worn by the owner of the pyxis or more likely a necklace of a childhood type that was saved in the pyxis as a memento. 13 The amulets depicted on the ex-votos from Paestum, the Vatican boy, and the

associated with the grave of a young woman contained yet another group of amulets that included an amber knucklebone held in place by a silver wire, two cast-silver double axes, and matching pairs of images in silver foil pressed from a mold: bees, turtles, frogs, trefoils, and the circular frontal heads of lions and silens.7 These shapes correspond well with some of those we saw earlier on the bandolier worn by the Vatican boy (Figure 1.6A),for example, the three double axes, the two trefoils, the open hand, and the dolphin. Other examples come from two late classical graves

Benaki pyxis are the most varied and detailed, and thus they allow us to make some useful comparisons with those from the fourth-century graves:14 Items from fourth-century

graves:

ABDERA (child): fica, scallop, double axe, dolphin, trefoil ACANTHUS: Heracles' club, cicada, scallop, double axe, open hand THASOS: (woman) knucklebone and pairs of double axes, lion heads,

in the Crimea. 8 The first (Figure 2.1) was found in the tomb of a woman, and it includes in the top row-from left to right-a bird, a scallop shell, and a double

bees, turtles, frogs, trefoils KERTCH (woman): scallop, double axe, peapod, animals, birds

axe.9 In the second row, it has a dolphin, another bird, and a shape that Reinach

KERTCH (woman): scallop, double axe, peapod, two birds, dolphin S. RUSSIA: Heracles' club, double axe, open hand, horizontal

called the "cosse de pois" ("peapod"), which also appears on the temple boy in Plate 3. The second necklace (Figure 2.2) is decorated with animals, two scallop

amulet case

shells, a double axe, and a "peapod." A third string found in southern Russia has a similar set of objects, 10 including Heracles' club, a horizontal amulet case, a dou-

Items depicted on fourth- to third-century images:

ble axe, and an open hand hung vertically. We know that the bone amulets from

BENAKI PYXIS (woman): two trefoils, open hand, dolphin, lunula,

girl. 11

Abdera belonged to a child of three years, probably a The first necklace from Kertch, on the other hand, was found in a woman's tomb, and we have other

cicada, double axe, single eye, cicada, peapod CYPRIOT VOTIVES: circles, lunulae, peapod, double-axe, horizontal

indications that women may have kept their childhood amulets into their adult lives and perhaps continued to wear them. A painting on a pyxis in the Benaki

amulet case VATICAN STATUE (boy): four spearheads, three double axes, three

Museum makes a similar suggestion (Figure 2.3). 12 Produced in a fourth-century BCE Attic workshop, it depicts on one side a necklace with fourteen amulets,

lunulae, two trefoils, open hand, and dolphin PAESTUM VOTIVES (boys): double axe, peapod, trefoils, lunula

58

Archaeology Among the six assemblages found in the graves we see two popular shapes: the scallop shell and the double axe, objects that perhaps represent a regional style,

armpit and neck. Nine were carved from amber (including a phallus, a fish, a hare, and a bird), and two others were shaped as traditional amulets: a tiny blue image

since all these graves are in the northern parts of the Greek world, from where the pyxis also may have come. 15 The types of amulets that appear on the images, how-

of the Egyptian protective deity Bes and a boar's tooth roughly carved in the shape of a crescent moon. 22 An infant in Vienne was placed in a stone sarcophagus wearing two silver bracelets with bells and a single amulet cord that held three pierced

ever, are somewhat different. The double axe, for example, belongs to this group, but the scallop never appears. In its place we find the ever present lunula. There are, moreover, five other objects that appear sporadically in both groups: the tre-

coins, five pierced shells, and fourteen pierced amber beads of birds, fish, cicada,

foil, the dolphin, the cicada, the open hand, and the "peapod," the last two of which could be suspended in either direction.

amber on some of these childhood strings is especially important because of its healing properties. Sometimes we find only a pair of shapes in a child's grave-for

A recent study of the classical and early Hellenistic graves in Apollonia Pontica reveals that thirteen children between the ages of six months and fifteen years

example, three different cases where a gold bulla was strung alongside a silver phallus, a club of Heracles, or a tiny statue of Isis-Fortuna, and a single case where a

old wore amulet strings on their torsos, wrists, or ankles, but of a more mundane or

small bronze bell is worn with a bronze phallus. 24 There is, finally, some evidence

homespun quality-for example, yellow or blue glass beads of the Phoenician or Punk type decorated with "eyes" or shaped like the heads of bearded men,

that young children wore single earrings as amulets. 25 Such strings of amulets are also found occasionally with the skeletons of young

seashells (often cowrie), pierced coins, and animal teeth. 16 Some were quite simple,

women who died fleeing the eruption of Vesuvius in 79 CE, perhaps providing us

like a single kouros-shaped bead carved of bone and suspended upside down on the neck, while others had many amulets, the most elaborate being a gilded silver

with a more realistic snapshot of daily amulet use, as opposed to the ceremonial

and once again the club of Heracles. 23 As we saw in Section 1.5, the presence of

medallion showing Peleus carrying off Thetis (mentioned earlier) with eight beads

..

(including four with eyes) and two animal teeth, all of which were probably worn

,t

on a bandolier string.17 The shorter strings were apparently worn on other parts of the body, for example, two children older than five wore shells and beads on

" ( l

their ankles, and two very young children wore beads, shells, and pierced coins on their right wrists. 18

)

//lk ·/

C

'

R

One important feature of these shells, glass beads, and metal coins is that they rattled when the child moved. Roman-period graves yield similar shapes and in some cases point to a continuous tradition, especially in Campania and further

59

Shapes

A

K

south on the Italian peninsula, where Greek influence was early and strong. The Latin word for such amulet cords is crepundia, derived from the verb crepere("to rattle" or "jingle"), and it is used by archaeologists to describe strings of small beads and amulets that are found mainly in children's graves. It is easy to assume, of course, that such crepundia, especially bracelets with bells, 19 were merely rattles or playthings given to cranky children to distract them, but often the media or shape of the individual items on these crepundia reveal that they were strings of amulets specifically designed to protect children from diseases and other dangers. 20 Phalli cast from metal, for example, as well as the animal teeth and pierced coins that we saw from Apollonia, show up frequently in the tombs of children, as on a bracelet from an infant's tomb that included two boar's teeth, a bell, and five pierced coins.21 And a one-year-old child buried in the suburbs of Rome had a set of seventeen pieces worn on a cord that was hung either on the child's neck or between its

A Ccuwp,,:, Vases B hw,terminal}u;un C . llan.d-1 ,:,vndnJl. 01'ls16· D , ·,Ta,,kc,7,,7,J:rrihnt;,,of'.Anribrh' E .Hr.aul m,okiny an1p )-.))...~\f.\)( V.J f'A> k (:,A

E "YE- ~'tOc't,....Ji' f',J\.L)V 0ri' I, )>. I ..o. 0.60:vo.f)),.o.

45

80.vo.P1..o. 0cpo:,:cuo:vaf)i ..o.

The proliferation of inscribed incantations of the Roman period, does not, then,

scribing on permanent media metrical charms that previously had been chanted

,,

;

T:VCUµO: O:)'tOV.

in much the same way that artisans in the same period encircle the frontal eye with powerful weapons and animals (Figures 4.1-3). 124 reveal any newfound superstition or need for magical protection; rather, this is a mirage created, as was the case with the prayers, by the newfound habit of in-

" ,

ic n:a:ci,p !:ii tC UtOC

v ap1..o. " Cl)'lE

C0:1£

Tcipo\:~,

xapax:-

Tiipcc

KCV

a

Tiai) ..0:8{0:,

arcoiCO:V'tOC piyouc, -rpr-ceov,1e~