The Unicorn in Muslim Traditions

= SMITHSONI AN INSTITUTIONS
FREER GALLERY OF ART

OCCASIONAL PAPERS
VOLUME ONE NUMBER THREE
Studies in Muslim Iconography
I. THE UNICORN
By
RICHARD ETTINGHAUSEN
WASHINGTON
1950
FREER GALLERY OF ART OCCASIONAL PAPERS
The Freer Gallery of Art Occasional Papers, to be published from
time to time, will present material pertaining to the cultures represented
in the Freer Collection, prepared by members of the Gallery staff.
Articles dealing with objects in the Freer Collection and involving

original research in Near Eastern or Far Eastern language sources
by scholars not associated with the Gallery may be considered for
publication.
The Freer Gallery of Art Occasional Papers will not be sold by
subscription, and the price of each number will be determined with
reference to the cost of publication.
=SMITHSQNI AN INSTITUTION ~
FREER GALLERY OF ART

OCCASIONAL PAPERS
VOLUME ONE NUMBER THREE
Studies in Muslim Iconography
I. THE UNICORN
By
RICHARD ETTINGHAUSEN
WASHINGTON
1950
Zt or& (gattimovz (press
BALTIMORE, MS., U. S. A.
FOREWORD
During his research the author benefited from information
given by manyof his friends and colleagues. He is indebted
to Prof. L.A. Mayer, Dr. Harold W. Glidden, and Dr. A. R.
Nykl for various valuable suggestions to Prof. Franz Rosen-
;
thal, who was always ready to put his wide knowledge of the
Islamic world at the writer's disposal and who was also kind
enough to read the manuscript before its publication; to
Dr. Schuyler Cammann for his many comments on Chinese
animal lore; and to Dr. Milton Anastos for information about
classical writers and especially for his translations of perti-
nent texts from the books of Aelian and Timothy of Gaza.
D. S. Rice kindly supplied him with the text and translation
s
of an unpublished passage in the Tab a al-hayawdn manu-
I
scripts in the India Office Library and the British Museum.

Zoological questions (which often must have sounded rather
queer) were given patient attention by Dr. Remington Kel-
logg, Director of the United States National Museum, and
Dr. Herbert Friedmann, Curator of the Division of Birds in
the same institution. References to other expert advice are
made in the text itself.
The procuring of proper illustrations was often difficult,
and had to be taken. The author herewith
special pictures
wishes to thank all owners and curators of miniatures, manu-
scripts, and art objects for their permission to have their
material included in this study. He is also indebted to Leo-
poldo Torres Balbas, Basil Gray, Ernst Kiihnel, Eustache de
Lorey, D. S. Rice, J. Sauvaget, Erich F. Schmidt, and J. V. S.
Wilkinson for supplying photographs from their collections,
many of which would have otherwise been unobtainable. Frau
Maria Sarre kindly permitted the reproduction of illustra-
tions in her husband's publications, which was all the more
gratifying since Professor Sarre was the first to describe one
of the motifs here investigated. Certain material reproduced
iii
iv Foreword
in thisvolume, namely, one plate which is specifically identi-

fiedby the use of captions appearing thereunder, was taken
from the Japanese work Shosdin Gyomotsu zuroku, vol. i
(July 1929) and vol. 7 (August 1934) published by the Im-
perial Household Museum, Tokyo. The Japanese interests
in the United States copyright in this work (12 volumes,
1 929-1940) were vested in 1949, pursuant to law. The use

of these plates in the present volume is by permission of the
Attorney General of the United States in the public interest
under License No. JA-1358.
All illustrative material was expertly processed by B. A.
Stubbs, of the Freer Gallery of Art. I am also indebted to
Mrs. Bertha M. Usilton, librarian of the Freer Gallery, for
checking and arranging the bibliography, and preparing the
index.
Finally the author wishes to thank his colleagues at the
Freer Gallery for their interest, and particularly the Director,
A. G. Wenley, for his continued patience.
The transliteration of Arabic used in this investigation

represents a simplified version of the system developed for
the Encyclopaedia of Islam. Common names such as Muham-
mad, Islam, Iran, Seljuk, and the like, or adjectives derived
from common names like Fatimid, Timurid, or Safavid, and
others are rendered in the usual English form.
A.D. dates precede the Hijra years which are given in
parentheses. When only one date is given, the Christian Era
is understood.
The literature cited in this paper is listed at the end of the
book. Each item is preceded by a bold-face numeral which
serves as a code number in references.
CONTENTS
Page
Foreword iii
List of plates vii
List of text illustrations xii
Introduction I
The iconographic problem 3

The identification of the animal 6
The physical appearance of the karkadann in Muslim texts and
illustrations 12
The karkadann attacking other animals 26
The hunt of the karkadann 35
The tamed karkadann 47
Additional features of the karkadann 52
Other unicorns in Muslim literature 59
The lore of the unicorn in the Muslim world compared with that of
other civilizations 73
A. The ancient Orient 73
B. The classical world 74
C. Biblical and early Christian writings. 92
D. Indian influences 94
E. Chinese influences 101
F. The late medieval world 109
The karkadann as a scientific and artistic problem 143
Literature cited 163
Index 193
V
LIST OF PLATES
(All plates at end of text.)
Plate
1. Glass vessel, gilded and enameled. Probably made for a Rasulid

sultan of Yemen. Syria, ca. 1330- 1340. Freer Gallery of Art,
No. 33-13-
2. Two details from the glass vessel shown on plate 1. Freer Gallery
of Art, No. 33.13.
3. Upper: Details of a silver-inlaid bronze canteen. Syria, first half
of the thirteenth century. Freer Gallery of Art, No. 41.10.

Middle: Stone sculpture from a portal: Unicorn (Karkadann)
pursuing an elephant. Konya (Anatolia), Seljuk period, middle
of the thirteenth century.
Lower: Stone sculpture from a portal: Unicorn (Karkadann)
pursuing an antelope. Konya, Seljuk period, middle of the thir-
teenth century.
4. Detail of an engraved and copper-inlaid bronze bucket. Iran,
twelfth century. New York, possession of R. Stora.
5. Left: Detail of an ivory box with the name of Hajib 'Abd
al-Malik b. al-Mansur. Made by 'Ubaida and Khair. Spain,
1005 (395)- Pamplona, Cathedral.
Right: Ivory box, related to Spanish boxes of ca. 970, but possibly
of later date. Paris, collection of the Marquis de Ganay.
6. Plaque from the portal of the Madrasa Muqaddamiya, Aleppo,
1 168 (564).
7. "Kardunn" from a Na 't al-hayawdn manuscript. Iraq, middle of
the thirteenth century. London, British Museum, OR. 2784.
8. Detail of a page from a Munis al-ahrdr manuscript, now dispersed.
Iran, 1341 (741). Cleveland Museum of Art.
9. "Iskandar Battling a Karkadan" from the "Demotte Shah-ndmah"

now dispersed. Iran, ca. 1320. Boston, Museum of Fine Arts.
10. "Karkadan" from a Manafi-i hayavdn manuscript. Iran (Ma-

ragha), end of the thirteenth century. New York, The Pierpont
Morgan Library, ms. M.500.
11. "Domestic Oxen" from a Mandfi-i hayavdn manuscript. Iran
(Maragha), end of the thirteenth century. New York, The Pier-
pont Morgan Library, ms. M.500.
vii
viii Freer Gallery of Art Occasional Papers Vol. i
Plate
12. Lohan on a rhinoceros, detail of the painting "Lohans Moving

Through Forests and Sea." Ink on paper. Chinese, Ming period.
Freer Gallery of Art, No. 19.174.
13. The karkadann in al-Qazwini manuscripts.
Upper, left: Iran, early fifteenth century. Berlin, Staatliche
Museen, Islamische Abteilung.
Upper, right: Iran, fifteenth century. New York, possession of
H. Kevorkian.
Lower: From the "Sarre manuscript." Iraq, late fourteenth cen-
tury. Private collection.
14. The karkadann in al-Qazwini manuscripts.
Upper: Iran, 1545 (952). London, collection of A. Chester
Beatty, Esq., ms. P. 212.
Lower: 1789 (the date given is that of the thirty-first year
India,
of the reign of Shah 'Alam). Freer Gallery of Art, No. 07.625.
15. Left: Island scene with a karkadann, from an al-Qazwini manu-
script. Iran, seventeenth-century adaption of a fifteenth-century
copy in a manuscript dated 1460 (865). Princeton University
Library (Robert Garrett collection, No. 82 G).
Right : The lure of the ring dove, from the same manuscript as
the left miniature.

16. Upper: "Shutur-gav-palang" from an al-Qazwini manuscript.
India, 1789. Freer Gallery of Art, No. 07.625.
Lower: "Sinad" from the "Sarre al-Qazwini manuscript." Iraq,
late fourteenth century. Private collection.
17. Right side of a (detached) double title-page from an al-Qazwini
manuscript. Iran, first half of the fifteenth century. New York,
Metropolitan Museum of Art.
18. Upper: The karkadann pursues the elephant. Luster tile. Iran,
thirteenth century. Berlin, Staatliche Museen, Islamische
Abteilung.
Middle: Detail of a silver-inlaid bronze plate, showing twice the
motif of the karkadann pursuing an elephant. Syria, first half
of the thirteenth century. Leningrad, Hermitage Museum.

Lower: Carpet in a miniature of an al-Hariri manuscript showing
elephant, rhinoceros, and 'anqa' (?). Egypt, 1337 (738). Ox-
ford, Bodleian Library, ms. Marsh 458.
19. Upper: The karkadann pursues an elephant. Detail of the silver-
inlaid vesselcommonly called Baptistere de St. Louis. Syria, ca.
1300. Paris, Musee du Louvre.

.
No. 3 List of Plates ix
Plate
Lower: Detail of a luster tile showing karkadann pursuing an ele-
phant. Iran, thirteenth century. Boston, Museum of Fine Arts.

20. Covered glass vessel, gilded and enameled. Syria, first half of the
fourteenth century. New York, possession of H. Kevorkian.
21. Mughal animal carpet. India, first half of the seventeenth century.
Washington, D. C, National Gallery of Art (Widener
collection)
c
22. The death of the elephant, in a manuscript of the Aja 'tb al-makh-
luqdt. Iran, late sixteenth or seventeenth century. Baltimore,
Walters Art Gallery, ms. W.593.
23. Isfandiyar killing two karg monsters, from a Shah-ndmah manu-
script written for Prince Baysunqur, son of Shah Rukh, dated
1429 (833). Teheran, Gulistan Palace Museum.
24. Upper: Gushtasp killing a karg, from a Shah-ndmah manuscript,
now dispersed. Iran, first half of the fourteenth century.
Lower: Iskandar killing a karg, from the same manuscript as
upper miniature.
Washington, D. C, collection of George Hewitt Myers.
25. Bahram Gur killing a karg, from a Shah-ndmah manuscript, now
dispersed. Iran, first half of the fourteenth century. Freer Gal-
lery of Art, No. 30.10, reverse.
26. Gushtasp killing a karg, from a Shah-ndmah manuscript. Iran,

second half of the fifteenth century. Paris, Bibliotheque Na-
tionale, ms. Supplement persan 1280.
27. Bahram Gur killing a karg, from a Shah-ndmah manuscript. Iran,
1497 (902). New York, possession of H. Kevorkian.

28. Isfandiyar killing karg monsters, from a Shah-ndmah manuscript.
Iran, 1497 (902). New York, possession of H. Kevorkian.
Iran, 1605-1608 (1014-1016). New York, Metropolitan Mu-
seum of Art, Cochran ms. No. 3.

Iran, 1544 (95 1 )- Princeton University Library (Garrett col-
lection, No. 56 G).
31. Left: Fragment of an unglazed pottery vessel. Mesopotamia,
thirteenth or possibly late twelfth century. Berlin, Staatliche
Right: Fight with a monster. Persepolis, Palace of Darius
{tachara) .
x Freer Gallery of Art Occasional Papers Vol. i
Plate
32. Babur on a rhinoceros hunt. Mughal India, second half of the

sixteenth century. Baltimore, Walters Art Gallery, ms. W. 596.
33. Jahangir on a rhinoceros hunt. Mughal India, first quarter of
the seventeenth century. Diisseldorf, O. Sohn-Rethel collection.
34. Bilqis surrounded by the animals of Sulaiman, from a Nizami

manuscript. Iran (Shiraz), 15 13 (919). New York, possession
of H. Kevorkian.
35. The host of Sulaiman. Left side of a double frontispiece, the
right side of which shows the enthroned Sulaiman and Bilqis.
From Shdh-namah manuscript, now dispersed. Iran, 1497- 1504

a
(902-910). Princeton, N. J., collection of Dr. E. Kahler.
36. Laila visiting Majniin. Mughal India, ca. 1600. New York, pos-
session of Nasli Heeramaneck.
37. The animal kingdom. Mughal India, early seventeenth century.
Freer Gallery of Art, No. 45.29.
38. The island of Waqwaq from an al-Qazwini manuscript, now dis-
persed. Iran, sixteenth century. Boston, Museum of Fine Arts

(formerly Goloubew collection).
39. Left: Leather girdle with plaques of rhinoceros horn and silver
buckle of a second girdle of the same type.
Right: Pendant ornaments in form of small containers (gosu) of
rhinoceros horn.
China, T'ang period, before A.D. 756, Nara, Shosoin.
40. Upper: Harish in form of snakes, from an al-Qazwini manu-
script. Iran, seventeenth century. Princeton University Library
(Garrett collection, No. 82 G).
Lower: The capture of the harish, from a Manafi-i hayavdn
manuscript. Iran (Maragha), end of the thirteenth century.
New York, The Pierpont Morgan Library, ms. M.500.
41. The harish in al-Qazwini manuscripts.
Upper: From the "Sarre manuscript." Iraq, late fourteenth cen-
Middle, left: Iran, early fifteenth century. Berlin, Staatliche
Middle, right: Iran, fifteenth century. New York, possession of
H. Kevorkian.
Lower: Iran, 1545 (952). London, collection of A. Chester
42. Upper: "Shadhahvar," from the "Sarre al-Qazwini manuscript."

Iraq, late fourteenth century. Private collection.
No. 3 List of Plates xi
Plate
Middle: "Sinad," from an al-Qazwini manuscript. Iran, seven-

teenth century. Princeton University Library (Garrett collec-
tion, No. 82 G).
Lower, left: "Shadhahvar," from an al-Qazwini manuscript. Iran,
fifteenth century. New York, possession of H. Kevorkian.
Lower, right: "Shadhahvar," from an al-Qazwini manuscript.
Iran, early fifteenth century. Berlin, Staatliche Museen, Isla-
mische Abteilung.
43. The lure of the shadhahvar's horn, in al-Qazwini manuscripts.
Upper: Iran, seventeenth century. Princton University Library
(Garrett collection, No. 82 G).
Lower: Iran, 1545 (952). London, collection of A. Chester
44. The mi'raj in al-Qazwini manuscripts.

Upper: Iran, early fifteenth century. Berlin, Staatliche Museen,
Islamische Abteilung.
Lower: From the "Sarre manuscript." Iraq, late fourteenth cen-
45. Drawing with a Chinese dragon (lung), a lion, and a Chinese
unicorn (ch'i-lin). Turkey, sixteenth century. Freer Gallery of
Art, No. 48.17.
46. "Yahmur," from a Mandfi'-i hayavdn manuscript, now dispersed.
Iran, ca. 1400. Freer Gallery of Art, No. 38.2.
47. Upper: Detail of lower horn shown in the middle picture, reveal-
ing the engraved naskhi inscription on the silver mounting.
Middle: "Unicorn horns." The upper consists of fossil bones in
a fifteenth-century mounting the lower is a narwhal tusk whose
;
outer surface has been scraped off for medicinal purposes. The
two end mountings of gilded silver date from the thirteenth
century, while the central mountings of bronze and the affixed
chain are from the fifteenth century. Venice, Treasury of San
Marco (since the fifteenth century).
Lower: Knife signed by Mihr 'Ali with handle of walrus ivory.
Iran, sixteenth to seventeenth century. Baltimore, Walters Art
Gallery.
48. Upper: "Hai-ma," detail on bronze mirror. China, T'ang period
(618-907). Freer Gallery of Art, No. 44.5.
Lower: "Asb-i abi" from a Manafi-i hayavan manuscript. Iran
( Maragha) end, of the thirteenth century. New York, The Pier-
pont Morgan Library, ms. M.500.
LIST OF TEXT ILLUSTRATIONS
Figure Page
1. Seal impression. Divine hero killing a winged bull. Mesopo-
tamia, ca. 1200 B.C. Formerly collection of Prof. Ernst
Herzfeld 44
2. Indian seal found in Tell Asmar. Middle of the third mil-
lennium B.C 83
3. Sharav (sarabha). Detail (drawing) from an embroidered
Indo-Portuguese bedspread. Goa, seventeenth century.
Brooklyn Museum 98
4. One-horned leonine monster from a Chinese bronze mirror.
T'ang period (618-907). Freer Gallery of Art, No. 29.17.
(Drawn by Mrs. Eleanor M. Jordan.) 105
5. Chinese mirror. T'ang period (618-907) 109
xii
STUDIES IN MUSLIM ICONOGRAPHY
I. THE UNICORN
By RICHARD ETTINGHAUSEN
Associate in Near Eastern Art
Freer Gallery of Art
[With 48 Plates]
INTRODUCTION
It is generally acknowledged that Islamic art is an art of
decoration ;
yet we have to admit that so far hardly any Mus-
lim sources have been tapped which explain the meaning and
mental associations of these decorative schemes. We do not
know, for instance, what a Muslim artist had in mind when
he painted an arabesque, a peacock, a hare, or the more fan-
tastic animals such as those which are usually called griffons
and harpies. Even the names of many designs are not known
to Western scholars. There is usually also no explanation to
be found as to why certain motifs became popular at certain
times and then disappeared.
The following study tries to establish the various icono-
graphic forms and the historical setting of the "unicorn"
motif. It also intends to reconstruct the connotations most
likely to be found in the mind of a medieval Muslim con-
fronted with a picture of the animal.
The unicorn is not a frequent theme, though it occurs more
often than one was hitherto inclined to believe. It is also
true that there are other figurative designs which are more
important. The varied iconographic uses found in a limited
number of representations provided, however, exceptionally
favorable conditions, first, for the identification of the motif,
and then for the interpretation of the various types. Finally,
the many connections of the motif with India, China, and the
classical and medieval worlds made it particularly attractive
1
2 Freer Gallery of Art Occasional Papers Vol. i
for a more extensive investigation. As a result not only has

the unicorn as such become more significant, but also it has
been possible to reconstruct some of the conditions and mental
processes which created the peculiar forms of the animal.
In all cases our witnesses were Muslim writers of the ninth
to the seventeenth centuries, ancient Asiatic and European au-
thors and, of course, actual representations of the animal in
its various forms found in Islamic countries and dating from
the tenth to the eighteenth centuries. The author readily ad-
mits that he shares the lot of all students of the lore of the
unicorn, namely, that even after a diligent search, there still
remains a certain amount of mystery. This applies especially

to the prehistory and genesis of the motif and to its odd
ramifications which, in many cases, could be presented only
as hypotheses. There is always the possibility that some un-
tapped source might throw further light on this fantastic ani-
mal. Yet, it is hoped that the main thoughts of the medieval
mind on this subject can now be grasped. A few of the illustra-
tions are admittedly of late date or of inferior artistic quality,
but they are the only traceable expressions of a particular
belief and are thus valuable as historical documents. Happily
the majority of the illustrations offered reveal the usual beauty
and finesse that we associate with Islamic art.
THE ICONOGRAPHIC PROBLEM
Among the fantastic animals which decorate objects and
buildings of the Islamic Middle Ages, one occasionally finds
a quadruped with a long horn on its forehead. An enameled
Syrian glass vessel of the fourteenth century in the Freer Gal-
lery of Art provides characteristic examples (pis. i and 2). 1
Along its outer rim, in the interstices between decorative
roundels, eight fantastic animals areshown against an ara-
besque background. Two are "griffins," two are "sphinxes,"
one is a winged lion, and three are winged "unicorns" with
feline bodies
one with a long straight horn and the two
others with a shorter curved one. Two earlier examples of
the monster are found in a less conspicuous place on a thir-
teenth-century inlaid canteen, also from Syria, in the same
museum (pi. 3, upper). 2
Here the winged "unicorns," which
cannot be defined from a zoological point of view, are to be
found together with many other fabulous and ordinary ani-
mals, and human beings as well, all set within a highly decora-
tive inscription on the side walls of the vessel. If there are
any doubts whether the animals are single-horned, or whether
they are meant to have two horns which only cover each other
in profile view, such doubts can be put to rest by the fact that
other animals in the same frieze are actually shown with two
horns.
Since the two objects in the Freer Gallery are both from
Syria, it should be pointed out at once that the motif is not
restricted to this particular country. Even a cursory survey
shows that it is also to be found in other regions. An engraved
and copper-inlaid bronze bucket of the twelfth century, from
Iran, in the possession of R. Stora, of New York, shows in
its central register between decorative quatrefoils a pair of
the animals portrayed in a heraldic style and posture (pi. 4).
1 No. 33.13. Cf., 160, vol. 1, pp. 404-405; vol. 2, pi. 179, No. 9.
2 No. 41.10. Cf., 83, fig. 3.
3
Here head and body are basically equine, although many

bodily features derive from other animals. Yet the long
straight horn puts the winged creature in the same category
as the Syrian examples. And just as in the case of the glass
vessel, the unicorns are grouped together in the same frieze
with other fantastic animals, viz, two griffins and four winged
3
lions.
From the opposite end of the Muslim world come a num-
ber of carved Hispano-Moresque ivory boxes which again
4
display the fabulous beast (pi. 5, left). In nearly all cases
the winged animal with its single horn showing helical grooves
is zoologically difficult to explain. It always faces another
specimen of the species and is set in a decorative scheme which
contains similarly stylized pairs of griffons. They are the
earliest representations of the unicorn in Muslim art so far
traced.
When we now turn to a fourth region, Anatolia, and look
this time for the use of the design within an architectural
framework, we can refer to unicorns mentioned by Friedrich
Sarre, who was one of the earliest, if not the first, to have
observed the motif. He described two examples found among
the decorative stone sculptures of Konya which apparently
came from an arched doorway (pi. 3, middle and lower). 5
These thirteenth-century Seljuk reliefs portray a different con-
ception inasmuch as they do not represent the animals as iso-
lated motifs as on the glass vessel, nor as pairs as on the
Persian bucket and the Spanish boxes, nor yet as a motif
within an over-all pattern as in the case of the Syrian metal
canteen. In both examples from Konya the winged feline
"unicorn" is part of a group; it is seen pursuing another ani-
mal, in one case an antelope and in the other an elephant.
Another example of the use of the motif as part of an
architectural decoration and in still another iconographic set-
ting is to be found on a plaque from the portal of the Madrasa
Muqaddamiya in Aleppo, which dates from 1168(564)
3
117, vol. 6, pi. 1291, B.
4
100, vol. 1, pis. 28, 33, 42, 48, 50, 56, and 59.
5
225, pp. io-ii, figs. 10 and 12; 227, pp. 53-54 and pi. 17.
No. 3 I. The Unicorn Ettinghausen 5
(pi. 6).
6
Two of its arabesque forms which are part of the
symmetrical "tree" arrangement terminate in what seem to
be antelope heads which carry a straight tapering and heli-
cally grooved horn on their foreheads. The horns are obvi-
ously meant to be single ones, because the artist showed in
his treatment of the ears on other heads that he was able to
give a profile view in proper perspective. There is, therefore,
little doubt that he would have been able to show a second
horn had he wanted to do so. At this stage of the investiga-

tion it is more difficult to answer another question, namely,
whether the helical grooves of the horn are just a fanciful
surface decoration or are meant to be a realistic rendering
of an actual horn. This is a question which will have to be
taken up later on.
All the cited examples showing the whole animal are winged
"unicorns." Since in many other respects they differ from each
other, however, the following questions arise: are these ani-
mals different unicorns, or are they variants of one and the
same animal, and what should we call them? The traditional
Western term "unicorn," descriptive as it is, does not give us
any clue as to what the medieval Muslim named the animal
and what he had in mind when he created such a monster or
what he thought when he saw it. The term "unicorn" is, in
a way, even misleading, since for a person steeped in the tradi-
tions of Western civilization the word has many connotations
which have been nurtured by classical, Biblical, early Christian,
and medieval beliefs about this animal. The motif has to be
studied and understood from the point of view of the medieval
Muslim. Only then can it be properly interpreted.
c I
owe the knowledge of this plaque and the photograph of it to the kind-
ness of M. Jean Sauvaget (235, vol. 1, p. 123, footnote 399). Ecochard thinks
that this medallion belongs to the most beautiful Arab decoration of the twelfth
century (91, p. 84, fig. 2, and pi. 6, fig. 1).
THE IDENTIFICATION OF THE ANIMAL

wants to establish the associations which were formed
If one
in the Muslim mind in connection with this fantastic beast,
one has first to establish its Arabic and Persian names. Only
after this initial step will it be possible to ransack Muslim
literature for the necessary information. This will also, in
due course, whether we have here in the
settle the question
several illustrations referred to, and in others to be adduced
later on, one and the same or different animals.
Happily a number of representations of a winged quad-
ruped with a single horn have come down to us, accompanied
by texts giving specific names. It will suffice to name three for
a proper identification. The earliest so far found is in a zoo-
logical manuscript compiled from the works of Aristotle and
of 'Ubaid Allah b. Jibril b. Bukhtishu' (Bokhtyeshu' ) the ,
Na ( t al-hayawan wa-mandfi'iihii in the British Museum, prob-

ably written and painted in Baghdad in the first half of the
thirteenth century. 1 In the section on wild quadrupeds one
finds an animal which, although it does not have the feline
character of the examples on the Freer glass bowl and the
Konya reliefs, or the more equine character of those on the
Stora bucket, nevertheless reveals the basic and distinguishing
features of the unicorn, the horn on the forehead and the
wings (pi. 7). In this manuscript this strange antelopelike
animal is called 5^ ^
kardunn, a variant of the more common
form J? karkadann, or
2
^
karkaddan, which is the
usual term in Arabic for rhinoceros. Another "unicorn"
1 OR. 2784, see 56, pp. 531-532, No. 778. For a bibliography of publications
on the miniatures in manuscript see 126,
this No. 33, and 63, p. 153, No. 33.
p. 14,
See also 62, pp. 34-35, and figs. 34-36.

2 There are many more names for the animal. First there are the variants
of the usual name, viz, JuS^ ^ karkand or karakand, fem. -U5^j karkanda
(99, p. 109, footnote; Ibrahim b. Wasif-Shah, in 98, vol. 1, p. 160, footnote 1;
and 77, vol. 2, p. 327), and jt-^\ karkadan (5, fol. 14b, see pi. 10, Persian
caption; and al-Idrisi in 98, vol. 1, p. 180; jl J? kazkazan, in the Cairo
6
this time with a feline body is to be found on a page of the

Persian Anthology entitled Miinis al-ahrdr fi daqaiq al-
ash'dr, dated 1341 (741), in the Cleveland Museum of Art. 8
edition of A If la'ila iva-laila, see 23, notes to vol. 3, p. 130, is obviously a textual
garble). To this group belong also the usual Persian designations: jji^S^
karkadan, j->i^ j kargadan, and <-5^
^ karg. jJ^ ^ , that is, without task-
did over the nun, is found also in the Arabic text of the "Sarre al-Qazwini"
manuscript. Then there are other names, some translations or loan words from
other languages and some probably used only in specific regions. Thus al-Jahiz
and others quoting him speak of ^X^\ jLoi^l al-himar al-hindi, "the Indian
ass," hereby using a term of Aristotle (145, vol. 7, p. 40; 146, vol. 7, p. 123;
6, fol. 134b; 7, fol. 88b; 201, vol.

(J J*^\ al-qarni
7, p. 315). The term <-i>Yl
al-anf, "the horny on the nose" (6, fol. 134b; 7, fol. 88b) is found only in
Marvazi, an author who quotes more classical information than any other
Muslim writer on the subject; it is, therefore, obviously an Arabic rendition
of the Greek 'pivotcepm. jU^Jj al-bishan (al-mu'lam) (marked or spotted)
bishan (and various other misread versions of this term) come from the
Sanskrit (see p. 94, footnote 91). ganda (45, p. 228), has the same origin.
jy khartut occurs in al-Akfani (277, p. 354) ;

while the form *<Z j>- or
li> U jS- khartit or khartlt is nowadays used in Egypt and the Sudan (see
p. 130). Al-Biriini and Marvazi state also that the African Negroes use the
term impila. This Marvazi explains as ,jf jplJ*-] Jin "buffalo cow" (6, fol.
135b). One wonders about a possible connection with the name of a certain
African antelope, the irnpala, which became confused with the rhinoceros. This
antelope is called mpala and nipala by the Barotse and Ngamiland natives,
'mpara by the Ovadirico, and umpara by the Makuba the Angolan or black- ;
faced impala is called ompala or ombara by the Ovambo (243, vol. 2, pp. 550
and 557).
There are also certain names which do not seem to occur in the classical
literature of Islam, viz, j Ji y \ abu qarn (father of the horn) (53, vol. 12,
p. 607), jjS umm qarn (mother of the horn), and oj^c- 'anaza all used
in the Sudan (174, p. 203). Other names are j zab'ari, j~j> mirmis,
and jA hirmis. The form j^l J^>-^ wahid al-qarn (unicorn), is obvi-
ously derived from Greek fiovoicepcos (37, p. 20; 174, p. 203).
Finally there are certain names which are occasionally given to the rhinoceros
without actually designating the same animal. They are jj j>- harish, and
^ll*" sinad. (A short survey giving most of the names is in 174, p. 203.)
Throughout this investigation the form karkadann will be used with the
exception of those cases in which a specific form in an Oriental text has to be
transliterated.
3
This miniature is still unpublished, but the manuscript is well known (126,
p. 19, No. 48; 63, p. 155, No. 48).
In a ribbonlike composition four animals are shown, one in

front of the other (pi. 8). The caption in the left-hand cor-
ner of the preceding line describes them as
^ slmurgh (the fabulous bird), J^v elephant, and jj^ J?
^ jf vulture,
'
karkadan. Our third example is in a miniature of the "De-

motte Shdh-ndmah" of about 1320 which shows a scene usu-
ally entitled "Iskandar Fights the Habash (i.e., Ethiopian)
monster" (pi. 9)/ This designation is correct; but while it
stresses the country where the animal occurs, the Persian cap-
tion within the miniature jjf jxSL*] pjj, "The Fight of
Iskandar with the Karkadan," states more explicitly what kind
of animal Alexander the Great is slaying. All three quoted
examples show differences in the body structure of the beast,
although each has the long single horn on the forehead and
the wings emanating from the sides of the body. It seems
therefore certain that the various winged creatures with one
horn are representations of the karkadann or rhinoceros quite
irrespective of whether the represented animal has the body of
a lion, horse, antelope or any other quadruped.
Now that the winged "unicorn" has been established as
being the karkadann, we can look for this animal in the illus-
trated scientific manuscripts such as al-Qazwini's 'Ajd'ib al-
makhluqdt or the various works connected with Ibn Bukh-

tishu'.
5
They will naturally provide us with additional
4
126, pp. 23-24, No. 64; 63, pp. 158-160, No. 64. The quoted caption is after
54, p. 106, No. 33.

5 The texts dealing with the karkadann in the Na't al-hayaivdn iva-mandji'uhu
in the BritishMuseum, OR. 2784, the Mandft al-hayaivdn in the Bibliotheque
Nationale, Arabe 2782, and the Tabd'i' al-hayaivan *va-khaivdssuhu in the
Princeton University Library, No. 203B (Garrett ras. Ar. 1065) are, disregard-
ing certain minor variations, identical. According to the Paris manuscript
these texts are extracts from a larger work entitled o jsy* Jf>\
t^-\
^sluJl Kitdb al-khaiodss mujarrib al-mandfi' (203, p. 501, No. 2782). In gen-
eral, the descriptions of the animals are said to be taken from a work called
jl j*Â Na't al-hayaivdn, which is attributed to Aristotle, while the dis-
cussions of their medicinal value come from the jl ^J>-] ^sL* Mandfi' al-
hayawdn of Ibn Bukhtishu'. Both were put together by an anonymous com-

piler (56, p. 531). A three-line postscript in 2, fol. 198a states, however, that
the karkadann does not occur in the Kitdb al-na't and that it is mentioned only
a
examples of this animal, since these manuscripts have been

illustrated for many centuries. The earliest karkadann in a
scientific manuscript found so far is the one in the British
Museum referred to above (pi. 7). The representation in the
Persian Manafi manuscript of The Pierpont Morgan Library
is only a few decades later and was painted in Maragha in
northwest Iran at the very end of the thirteenth century

6
(pi. 10). The most striking feature of this animal when
compared with the representations so far discussed is not so
much its bovine appearance, but rather the fact that it lacks
the wings. Nevertheless, it is called a karkadan in the cap-
tion and illustrates a chapter enumerating all the traditional
features of the animal. We can make the same observation
for nearly all the other illustrations of the karkadann in
scientific manuscripts from this period on. It therefore be-
comes obvious that the wings in contrast to the horn are
not an essential feature of the karkadann. But we shall have
more to say about the wings later on.
Miniatures illustrating the al-Qazwinl text are to be found
for instance in the "Sarre manuscript" of the second half of
the fourteenth century (pi. 13, lower), 7 a derivative copy of
this manuscript in the A. Chester Beatty collection in London
(probably of the fifteenth century) 8 two other fifteenth-century

,
manuscripts the earlier one in the Berlin Museum (pi. 13,

upper left), 9 and the other from later in the century in the
Kevorkian collection in New York (pi. 13, upper right) 10

in the work of Ibn Bukhtishu'. The Arabic
text in the London, Paris, and
Princeton manuscripts is from the Persian Mandfi'-i hayavdn
totally different
(Pierpont Morgan Library, ms. M.500), which obviously goes back to another
Arabic model from which it was translated about 1295 (695). In this study
the two texts are referred to as Ibn Bukhtishu' and Mandfi'-i hayavdn (or, for
short, Mandfi').
6
126, p. 22, No. 57; 63, pp. 156-157, No. 57. Laufer was the first to identify
the animal by reading the Persian caption. Anet before him had called it a
"horned gnu" (165, pp. 88-89).
7
230, p. 38, No. 295.
8
Ms. P. 127, unpublished.
9
157, p. 69, fol. 272a (unpublished miniature). Kuhnel dates the manuscript
in the late fourteenth century which, I think, may be slightly too early.
10 Unpublished.
io Freer Gallery of Art Occasional Papers Vol. i
manuscript dated 1545 (952) in the Beatty collection (pi. 14,

upper), 11 another of 1570 (978) in the Harvard College
Library, 12 then what seems to be a seventeenth-century adapta-
tion of a fifteenth-century manuscript in the Princeton Uni-
13
versity Library (pi. 15), a Turkish translation of the text,
written in 1709 (1121) in the Walters Art Gallery, 14 and
finally, an Indian copy of 1789 (corresponding to the thirty-
first year of the reign of Shah 'Alam) in the Greer Gallery

of Art (pi. 14, lower). 15 Most of these seem to be different
from each Only a few show, at first glance, that they
other.
follow same iconographic tradition; for instance, the
the
British Museum and the Kevorkian examples (pi. 7 and pi. 13,
upper right) show an antelopelike animal, but only the uni-
corn in the older manuscript is winged. If further proof were
necessary to show the lack of uniformity in the conception of
the karkadann, it could be demonstrated by the different ap-
pearances of the same animal in a single manuscript. The al-
Qazwini in the Princeton University Library shows the ani-

mal by itself, and then again in its eastern habitat together
with other creatures (pi. 15, right and left). No uninitiated
person would ever think that the artist meant to illustrate
one and the same animal; nor is this a unique case of the
dual personality of the karkadann, because the al-Qazwini
manuscript Harvard College Library shows the same
in the
phenomenon. However, these divergencies do not seem to

have perturbed the painters or the reading public.
How can this iconographic variety be explained?
The step in finding an answer to this question is to con-
first
medieval texts which give the various conceptions of

sult the
the body structure of a karkadann. They also help us to
11 No.
43, pp. 139-140, No. 176, pis. 96-97; 115, p. 60, 19.
12 Unpublished.
13 106, p. 31, No. 65, states that this manuscript was copied in 1460 (865)
by 'Abd Allah b. Ali Bey Damavandi. Since the miniatures seem to be later
copies of fifteenth-century originals, the manuscript was either copied in toto
in the seventeenth (or even eighteenth) century or the miniatures are later
additions to the fifteenth-century text.
14 Ms. W.659, unpublished.
15 No. 07.625, unpublished.
No. 3 I. The Unicorn Ettinghausen
understand why
was represented in Muslim art at
the animal
all. The following quoted and paraphrased references are
culled from the published and unpublished writings of a large
number of writers of the ninth to seventeenth centuries, com-
prising litterateurs, historians, travelers, cosmographers, phar-
macologists, poets, merchants, and hunters. 10 Since it is the
aim of paper to explain a figural motif in Muslim art and
this
not to write a monograph on the rhinoceros, only the more
important passages which explain the outer form and the
meaning of this motif have been used while other information
is usually disregarded even though it may have literary or
folkloristic significance. For similar reasons discussions of an

etymological nature have also been left out. It is certain that
the list of authors referred to could be further enlarged, yet
there seems little doubt that the most common conceptions of
the "unicorn" within the Muslim world have been established
here and that they suffice to explain the paramount features
of the motif.
16 Following is a list of the most important authors and anonymous books most
frequently quoted in this paper. If the year of completion of a text referred to
is known, it is listed; otherwise the year during which the author died is
given, with the Hijra year in parentheses: Ibn Khordadhbeh, ca. 846 (232) ;
Akhbar al-Sin wa'l-Hind (formerly attributed to Sulaiman, the merchant),

85 1 (
2 37) J
al-Jahiz, died 869 (255) ; Ibn Fadlan, ca. 922 (310) ;
al-Mas'udi,
956 (345); Hudud al-Alam, 982 (372); al-Nadim, 987 (377); Abu Haiyan
al-Tawhidi, died after 1010 (400) al-Birunl, died 1048 (440) 'Ubaid Allah
; ;
ibn Jibril ibn Bukhtishii', died after 1058 (450) Ibn al-Balkhi, beginning ;
twelfth century (seventh century) Marvazi, ca. 1120 (617) Abu Ham id; ;
al-Gharnati, 1169 (565) al-Qazwini, died 1283 (682) Mandfi'-i hayavan, before
; ;
!295 (695) al-Dimashql, died 1327 (727)

;
al-Nuwairi died 1332 (732) ;
;
Hamd Allah al-Mustawfi al-Qazwini, 1339 (740) Ibn Battuta, before 1356 ;
(757); al-Damiri, 1371 (773); Ibn al-Wardi, ca. 1446 (850); Babur, died
I 53 (937) Abu '1-Fadl 'Allami, died 1602 (1011) and the A If laila iva-laila.
I ;
If a passage on the rhinoceros occurs only in one place in the Arabic or

Persian source, it is quoted only the first time it occurs in the following text.
THE PHYSICAL APPEARANCE OF THE KARKADANN IN
MUSLIM TEXTS AND ILLUSTRATIONS
Of all medieval authors al-Biruni alone in his book on India
gives a fairly accurate description of the rhinoceros: "It is of
the build of a buffalo, has a black, scaly skin, a dewlap hang-
ing down under the chin. It has three yellow hooves on each
foot, the biggest one forward, the others on both sides. The
tail is not long. The eyes lie low, farther down the cheek than
is the case with all other animals. On the top of the nose
1
there is a single horn which is bent upwards . .
."
The correct position of the horn and the unique observa-

tion about the hoofs indicate that this description is based on
good observation. Al-Biruni was not, however, the only Mus-
lim writer who had a chance to watch the animal at close
range. Ibn Battuta saw a rhinoceros in its Indian habitat on
four different occasions. What struck the Moroccan traveler
2
in particular and induced him to speak of it in two places was
the excessive size of the coarse head. He believed it to be

several times larger than that of the elephant, although he
pp. 203-204. Marvazi has another version of

1
45> PP- 229-252; 44, vol. 1,
al-Biruni's account which is more than twice as long and thus even more
specific (6, fol. 135a; 7, fol. 89a-b). Since certain passages are not quite clear,
we are quoting it here only in the footnote: "It is of the build of the buffalo,
but taller and short in the leg. The skin is smooth, not hairy, but scaly, made
of scales raised from and the chest (?) with dewlaps on both
the epidermis
cheeks. The haunches (of the animal) are big and the head flat-nosed and
receding. The horn is on the tip of the nose, conical in shape, and bent back-
ward toward the head. It is longer than a span. In the center of the upper
lip, under the horn, it has something like the additional "finger" on an ele-
phant's trunk. The lower jaw is like that of the bull; it has two blunt canines
inside the mouth. The nose resembles most closely the noses of beasts of burden.
The ears protrude on both sides, like the ears of the donkey. Its eyes are
almond-shaped (??) and set lower than is usual. Its tail is short, thick at the
root then widening toward the end. Testes and penis are like those of the
bull. Its hoofs are fleshy and resemble the feet of elephants; each has three
toenails, white with a yellow tinge; the biggest in front, then two in a half-
circle to the right and left."
2 and
I33 vol. 3, pp. 100-101, 356.
12
describes the rhinoceros correctly as being the smaller of the

two animals. In this instance it is possible to trace the source
of the misconception to which Ibn Battuta himself refers, i.e.,
to a popular saying "Karkadann head without body,"

(jJo ^ crb j-^Ôl). The reports of this famous traveler
thus show that even repeated encounters with the animal did
not necessarily produce a description based on observation.
Preconceived ideas prevented the eye from seeing reality.
In comparison with al-Biruni nearly all other authors had
only a vague idea of the karkadann. This lack of substance
led, of course, more and more to the invention of imaginary
traits. Al-Jahiz in the ninth century is still in the realm of
reality when he classes it with the largest animals such as the

elephant and the buffalo; 3
but at the time of Abu Hamid al-
Gharnati in the twelfth century, the rhinoceros has become

about 100 cubits long. 4
The karkadann is usually defined by comparing it with
other animals. A typical description in this case based on

accounts of the Volga Bulgars- is given about 922 by Ibn
Fadlan: "The animal is smaller than the camel and bigger
than the bull. Its head is that of a camel, its tail that of a
bull and its body that of a mule. Its feet are like the hoofs
of the bull. On
head it has a single, thick
the middle of its
and round horn which becomes thinner toward the tip " 5 . . .
When one examines the various comparisons, frequent analogies

with certain animals become apparent. The most common
statement is that the karkadann resembles a buffalo or a bo-
vine animal, although it is in most cases believed to be bigger,
once even larger than several buffaloes put together. 6 Al-
Damiri alone at one point gives the size as being smaller than
that of a buffalo. The Mandfi'-i hayavdn mentions, too, that
3
145, vol. 7, p. 29.
4
99, Arabic text p. 109, translation p. 266. This measurement is repeated
by al-Damiri, who speaks of "one hundred cubits and even more than that"
(77, vol. 2, p. 327).
5
134, Arabic text p. 33, translation p. 76.
6
5, fol. 15a, line 1, Abu M-Fadl 'Allami states that the rhinoceros is twice
the size of a buffalo (24, vol. 3, p. 120), while Babur gives the figure as
approximately three times the size (32, vol. 2, p. 489).
the tail and hoofs of the karkadann are similar to those of

the buffalo.
7
The '
'Ajaib al-makhluqdt manuscript in the Wal-

ters Art Gallery 8 likewise maintains that not only the parts
of the body just referred to, but also the horn, are those of
a bull. According to the text of al-Qazwini edited by Wusten-
9
feld, it has the constitution of a bull, only more powerful,
while Mustawfi thinks that it has the form of a cow. 10 From
these recurring comparative descriptions, it seems natural
that many representations show a bovine animal resembling
a buffalo or ox. Examples of this iconographic type are to
be found in the Morgan Mandfi' and the al-Qazwini manu-
scripts of the Sarre and Beatty collections and of the Prince-
ton University and Harvard College Libraries (pis. 10; 13,
lower; and 15, right). The Morgan Manafi' manuscript, like
many later zoological manuscripts, presents a welcome oppor-
tunity for comparing the miniatures of the karkadann and the
domestic ox (j^) which, as usual in Persian miniatures, is
represented as a zebu with its characteristic hump and large
dewlap (pi. 1 1 ) If one disregards such features as the curved
.
horn on the forehead, the different form of the muzzle and the
zebralike stripes, the karkadann appears as a bigger and
fiercer cousin of the domestic cow. The reason for the differ-
ence between the two will become obvious later on when it
will be shown that there are also other factors besides the
descriptive texts which have to be considered when
Muslim
one wants to explain the special character of an iconographic
type.
Of special interest is a miniature in the Beatty al-Qazwini
manuscript of 1545 (pi. 14, upper). In spite of the bovine
look, especially of its head and the hump on the back, the
karkadann also has certain elephantine features. It may be
thought that this is due to the ineptitude of the artist in por-
7 fol. line 8.
5, 15,
8 1, fol. 188. The title of this work is the same as that of the famous work
by al-Qazwini, but the text is quite different from the edition of Wiistenfeld.
The manuscript was written in the late sixteenth or the seventeenth century
for one Shams al-din Muhammad b. Mahmud b. Ahmad al-Salmani al-Tusi.
9
217, vol. 1, p. 402.
10 191, Persian text translation 25.
p. 35, p.
.

5
traying a rhinoceros, yet the medieval texts reveal certain

clues as to why the animal was given this form." While Ibn
Khordadhbeh states only that the animal is smaller than an
elephant and larger than a buffalo, 12 other authors go beyond
a Al-Qazwini, whose text the miniature
mere comparison.
illustrates, says definitely that the karkadann has the body of
an elephant, and according to al-Damiri it is a cross between
a horse and an elephant. 1 "
Further relationship between the
karkadann and elephant is brought about by the myth of the
11
jointless legs, a feature sometimes associated with the ele-
15
phant, other times with the karkadann. Still another con-
nection between the two animals is provided by the accounts
of the strange gestation of the female karkadann. As far as
we know, was al-Jahiz who introduced the story into Muslim
it
literature, though not without some misgivings as to its ve-

16
racity. According to him, shortly before the time of birth
the karkadann foetus sticks out its head from inside its
11
Another manuscript that shows the animal with elephantine features, such
as elephant legs and tail, is the late al-Qazwini manuscript of 1789 in the Freer
Gallery of Art (10, fol. 465a); here the rhinoceros also wears a saddlecloth
as does the tame riding elephant (pi. 14, lower; see also p. 48, footnote 6 of this
study)
12 This statement for instance, found also al-ldrisl
98, vol. 1, p. 25. is, in
(98, vol. 1, p. 180).

13 This is in line with the medieval Muslim concept that certain animals
were actually hybrids of well-known creatures. "Man allying himself with

the panther, the hyena, and other beasts has begotten the monkey, the
nasnds, and other creatures which resemble him. And thus it is that the
mingling of pigs and buffaloes has produced the elephant, of dogs and goats,
the wild boar ..." (65, p. 34). Al-Jahiz and al-Qazwini tell us that the
giraffe is believed to be the hybrid of a wild she-camel, a wild cow, and a male
hyena (146, vol. 1, pp. 142-143, 217, vol. 1, p. 383), or of the camel and the
panther, according to al-Mas'udi (184, vol. 3, p. 3).
14 The Akhbdr al-Sln iva'l-Hind see also 18, vol.
(17, p. 14; 1, p. 29;
vol. 2, p. V\ ) states: ". . . the karkadann has no joints, neither at the knees,
nor at the feet, since it is just one piece of flesh from the foot to the armpit."
Al-Mas'udi has another angle on the subject: "... most of its bones seem
to be grown together without joints in the legs, so that it cannot kneel or
sleep lying down, but has to lean between trees in the middle of the jungle
when it wants to sleep" (184, vol. 1, pp. 385-386).
15 See the discussion of this relationship on pp. 99-101 of this study.
10 see also pp. 143 of this study. The story also given
145, vol. 7, pp. 40-41 ; f. is
by Marvazi, al-Gharnati (99, p. 101, translation p. 267), al-Damiri (twice),

and al-Ibshihi (99, p. 267, n. 1).
1 6 Freer Gallery of Art Occasional Papers Vol. i
mother's womb and eats branches of trees. When it is satis-

fied, it pulls its head back into its usual place. In spite
17
of certain doubts raised against it, this belief continued
throughout the Middle Ages. There are, however, certain
authors like al-Qazwini and al-Damiri who transfer the story
of the karkadann foetus to another animal called sinad
18
(jl^) which, to quote al-Qazwini, "is of the same descrip-
tion as the elephant, only smaller in body." This elephant-
like sinad seems to have given further impetus to the mental
associations between karkadann and elephant. No miniatures
have so far been found which show the karkadann with its
enterprising foetus, but there is a miniature of the sinad in
the al-Qazwini manuscript which formerly belonged to Pro-
fessor Sarre, and others much like it are in the collections of
Museum and of Mr. Kevorkian (pi.
the Berlin 16, lower).
All show the dam with its young one partly outside the
womb and looking for supplementary food, and in each case
the sinad looks exactly like an elephant There is, however,
proof that the Muslim mind retained a connection between
the sinad and the karkadann. A crude Turkish al-Qazwini
manuscript of the year 1709 (1121) in the Walters Art
Gallery shows, as illustration to the chapter on the sinad, an
elephant with hungry foetus well visible in the back. 19 The
its
mother animal shows a curved horn on its forehead, although

the al-Qazwini text does not mention this feature; there seems,
therefore, little doubt that it was influenced by the concept of
the horn of the karkadann. All these weird associations of
the karkadann with the elephant give a clue as to why a
Muslim artist who had never seen a rhinoceros could have
been induced to give it elephantine features.
References to a resemblance of the karkadann to the horse
are rare in earlier texts, but the belief can be traced through-
1, pp. 387-388. See also pp. 62 f. and 144

17 184, vol. of this study.
18 This is the vocalization of the "Sarre manuscript" of the al-Qazwini text.
The same spelling is also given in 191, translation p. 18, and al-Damiri (78,
vol. 2, part 1, The Turkish al-Qazwini manuscript
p. 81). of 1709 (1121)
of the Walters Art Gallery (9, fol. 104a) has û- sanndd.
19 fol. 104a.
9,
No. 3 I. The Unicorn Ettinghausen n
out Muslim literature. Marvazi compares its size to that of
a horse, while the 'Ajd'ib al-makhluqdt manuscript in the
Walters Art Gallery mentions that the karkadann has an
equine neck. 20 Al-Damhi speaks of the animal as a hybrid of
elephant and horse, an idea which goes back at least to the
first half of the twelfth century, since such a pedigree plays
a decisive role in al-Zamakhshari's discussion about the law-
21
fulness of the karkadann's meat. The most pertinent state-
ment comes, however, from the Mughal emperor Babur, a
man who had seen many rhinoceroses in his life; yet accord-
ing to him "the rhinoceros resembles the horse more than it
does any other animal," a theory which he backs up with a
number of anatomical features which are said to be common
22
to both animals. Abu '1-Fadl 'Allami, another Indian writer,
makes the imaginative remark that the rhinoceros "much re-
sembles a horse in armor" however, it seems that these two
;
23
Indian authors are only rationalizing a popular myth. As
in the case of other comparisons this belief in a resemblance to
the horse had its iconographic repercussions and we can thus
account for some of the horselike features in certain repre-
sentations (pis. 1 8, upper; 19; and 20).
The affinity of the karkadann to the lion, or at least to parts
of its body, can be explained by some passages in the Shdh-
ndmah which speak of the lion claws of the karg <S (which
24
is one of the Persian words for rhinoceros). Of a later date
is the statement of the 'Ajaib al-makhluqdt in the Walters
Art Gallery that the karkadann has the head and feet of a
lion. This characterization fits the pictures on the enamel
glass bowl in the Freer Gallery (pis. 1 and 2) and of the
monster killed by Iskandar in the Demotte Shdh-ndmah
25
(pi. 9). Claws of a lion are also attributed by al-Jawhaii
20 188a.
1, fol.
21 Quoted in 77, vol. 2, p. 327, line 24.
22
32, vol. 2, p. 490.
23 120.
24, vol. 3, p.
24 102, vol. 4, pp. 314-315, line 423; 103, vol. 4, p. 339.
25
Another example of a lion-bodied unicorn is to be found among the stone
sculptures of Kubatchi in Daghestan. One of these twelfth- to thirteenth-century
reliefs shows two heraldically rendered lion-unicorns standing on their hind
i8 Freer Gallery of Art Occasional Papers Vol. i
26
to the harish (h rysh J^j>- ) >
an animal which this author,
Ibn Bukhtlshu', and al-Damiri identify with the karkadann.
(Other writers such as al-Tawhldi, 27 al-Qazwini, al-Mustawfi,
and the Mandfi' differentiate, however, between the karkadann
and the harish, which they regard as "a small animal the
size of a kid." This is obviously another tradition which will
have to be dealt with in the discussion of the harish later on
in this paper.)
Still another iconographic type is the karkadann in the
shape of an antelope or stag 28 such as we find in the British

Museum Ibn Bukhtlshu' and the Kevorkian al-Qazwini manu-
scripts (pis. 7 and 13, upper right). Its origin is rather com-
plex. First it can be said that the shape suggests itself to the
imaginative mind on account of the analogy with either ante-
lopes, the usual horned animals, or with stags, the bearers
of antlers. This iscorroborated by two passages in the Shdh-
namah which Firdawsi speaks of a terrorizing monster,
in
obviously a rhinoceros, which has a horn like the antler of a
29
stag. Another lead is the text of Ibn Bukhtlshu' which de-
clares the karkadann and the kidlike harish to be identical.
How this kidlike creature was then changed into an antelope
can be explained only after more material has been presented
in this paper (see pp. 96 f. of this study) But it seems obvious .
legs; the straight horns cross each other (36, pp. 95 and 113, pi. 68). This
Caucasian relief is not regarded as Muslim work, but it displays strong Muslim
influence.
26 The passage from the Sihdh quoted in 77,
is vol. 1, p. 285 ; 78, vol. 1,
p. 525.
-'Quoted in 77, vol. 1, p. 285; 78, vol. 1, p. 525.
2SEven for a zoologist it is often difficult to decide whether certain animals
in Muslim art are antelopes or stags, and pictures are of no avail in deter-
mining whether the animal drops its antlers annually as the stag does, or
keeps its horns permanently like the bovine antelope. Since the latter are much
more common than stags in Muslim regions, we usually speak of antelopes in
this investigation in spite of Firdawsi's poetic comparison of the horn of the
rhinoceros with the antler of a stag.
29
102, vol. 4, pp. 312-313, line 399; 103, vol. 4, p. 338; and 102, vol. 4, pp.
494-495, line 1605; 103, vol. 5, p. 122. In both cases the horn jj^, 3j~>, (sarun,
suru) is said to be like that of the j^'j^" gavaznan, which Wolff (280,
p. 738) translates as "Hirsche."

No. 3 I. The Unicorn Ettinghausen [9
that the confusion with the harish first found in Ibn Bukht-
ishus text was a decisive factor and it is therefore not sur-
prising that the first instance of an antelopelike karkadann
occurs in a manuscript of this author. Once the antelope type
was created by Firdawsi, Ibn Bukhtishu', and the additional
outside influence about which we have to speak later on, it
was then transferred to al-Qazwini manuscripts, such as the
one in the Kevorkian collection, although this particular author
does not confuse the karkadann with the imaginary harish.
A unique iconographic variety, and a very strange one at
that, is the animal in the Berlin al-Qazwini manuscript which,
in spite of two horns and four excrescences on its back, is
captioned as karkadann (pi. 13, upper left). The pertinent
texts, whether they are in al-Qazwini himself or in other
authors, fail to provide the clue for this particular shape. It
can however be explained as a substitute for another animal

with which it was confused since it shared certain features.
In his book on India, al-Birunl tells us of a strange two-
horned monster, larger than a rhinoceros and of the shape of
30
a buffalo, called sharav ( i^). It has four feet and in addi-
30
45, pp. 228 and 251; 44, vol. i, p. 203. Al-Biruni's passage is quoted by
Awfi, who died in the second quarter of the thirteenth century (93, vol. 2,
pp. 203-204; and 198, pp. 37 and 257). Al-Biruni and 'Awfi in turn are quoted
by al-Mustawfi (191, p. 43, translation, pp. 30-31) where for reasons which
we shall discuss on pp. 33 f. the animal is called rukh ^j. This shows that the
Indian lore about the animal is to be found in Muslim literature from the
eleventh till at least the middle of the fourteenth century, which explains its
occurrence in an early fifteenth-century manuscript.

There seems to be no consistent rule as to how the name 3 j> should be
pronounced, since it may never have been more than a bookish name in al-BIruni
and his followers. Reinaud rendered it as "scharau."' This form can also be read
sharav which is closer to Sanskrit s'arabha and which Reinaud recognized as
the source for the Arabic word (4s, p. 251). Sachau's transliteration of sharava

seems as far as the last vowel is concerned hardly possible in Arabic.
Muhammad Nizamu sharw and sheru, while Elliot gives
'd-din used the forms
sharu. Although it would be more correct to use shrw through this study, I
decided on sharav for the sake of readability and because it is closest to the
Sanskrit. According to the Sanskrit dictionary the sarabha is a kind of deer,
or, in later times, a fabulous animal supposed to have eight legs and to inhabit
snowy mountains; it is stronger than the lion and the elephant (188, p. 1057,
col. B). This eight-legged monster occurs in the Mahabharata (127, p. 18).
tion, on its back, something like four feet directed upward.

The four excrescences on its back and the two horns of the
animal in the Berlin manuscript demonstrate that this karka-
dann is actually a sharav. What has led to this substitution
will become obvious when more of the lore of the karkadann
has been investigated.
All reports pay special attention to the horn of the karka-
dann. Nearly all of them go back to confused hearsay or the
wrong interpretation of actual horns. Sometimes the accounts
take the medieval form of quoting an old authority. Al-
Jahiz, for instance, refers to Aristotle who "assumed that the
31
animal had one horn in the middle of its forehead." An-
other early account, that of the Akhbdr al-Sln wa 'l-Hind
reports too that the animal's single horn was on its forehead.
Ibn Bukhtishu', Asadi, Marvazi, and the Mandfi' again speak
of a single horn on the forehead of the animal, 32 while other
authors avoid giving specific information on this point. As we
have pointed out above (p. 12), al-Biruni was the only medi-
eval writer to describe its correct position.
Dimensions, form, and character of the horn also come up
for some discussion. Al-Jahiz thought that itwas harder than
the tusk of the elephant, thicker than a cubit, and its length
33
corresponding to its thickness. Ibn Bukhtishu' gives the di-
mensions as 1 cubit long and 2 closed-hands thick. The

Mandfi' says that it is 1 cubit long and adds that it is stronger
than the tusk of the elephant, more pointed and harder, all
of which makes the blow very effective. There is another vivid
statement in al-Idrisi, who tells us that the long horn is so
34
thick that one cannot encircle it with two hands. In some
31
145, vol. 7, pp. 40 and 42 (here without reference to Aristotle). He adds
also as a further authoritative support: "On this fact karkadann
(that the
has one horn in the middle of its forehead) the Indians, the old and the
young ones, agree."
32 There is an extinct Siberian rhinoceros that had a single horn on its
forehead (170, p. 26), but it seems most unlikely that this fossil animal
inspired the myth.
33
145, vol. 7, p. 38.
34 98, vol. p. 181. This is an exaggeration of the earlier statement by
2,
Ibn Khordadhbeh that the horn was two palms thick (98, vol. 1, p. 29).
editions of the Arabian Nights one can see popular imagina-

tion at work because in certain manuscripts the copyist was
not satisfied with the usual and the measurement is i cubit
35
therefore enlarged ten times. Al-Damiri does not give the
exact measurements but concludes that on account of the
weight of the horn, the karkadann is not able to raise its
head. A rather unusual at first sight even fantastic
feature
is first mentioned by al-Qazwini: ". on its horn is a . .
branch or protuberance (L*z sluiba) whose curve runs coun-

ter to the horn and this protuberance possesses special prop-
erties. The sign of its genuineness is that in it one sees the
figure of a horseman. One finds this protuberance only in the

possession of the kings of India." Al-Mustawfi, too, speaks
of this branch or protuberance which, according to him, is one
36
cubit (jT) long and has knots {zy> 'uqud). Al-Damiri re-
peats the text of al-Qazwini nearly word for word and he

places the branch on the tip of the horn. The possible mean-
ing of this curious development of the horn deserves special
discussion which will have to be reserved for a later section.
Only a few writers make any reference to the curve in the
horn. Al-Biruni is the first being followed by
to observe it,
the Mandfi' . Al-Qazwini even defines the character of the

37
curve. It is strange that Ibn Battuta, otherwise such an
astute observer, does not mention this characteristic feature,
just as he was unable to realize the peculiar position of the
horn. Yet he must have seen it at close range when, in two
instances, the severed head of the animal was brought back by
hunters. His only observation is that the karkadann "has be-
tween the eyes a single horn which is about three cubits long
and about one span thick." 38 This shows how difficult it was
35 comment The
See of 23, vol. 3, p. 386, footnote 36. Calcutta edition trans-
lated by Littmann speaks of a horn of 10 cubits (21, vol. 4, p. 127).
3(5
Al-Qazwini mentions in a story the knot of the horn j-\5^ ^50 | j^s jJit-
'uqdat qarn al-karkadann, without specifically stating from what part of the
horn it comes.
37 "In it is a curve which is convex towards its face and concave towards its
back."
38 100.
133, vol. 3, p.
3
to get a true picture of the animal and how much time had
to elapse before a realistic representation appeared.
The two-horned varieties of the karkadann, the one from
Africa and the other from Sumatra, are only occasionally
mentioned. 39 Al-Gharnati, with his predilection for the won-
40
derful, goes so far as to speak of a three-horned karkadann
which, according to him, has "one horn between its eyes and
41
two above the ears."
The rare reference to a curve in the horn is apparently one
of the explanations why only a few illustrations show it. The
outstanding example is the miniature in the Morgan Mandfi'
where the horn is placed on the forehead of the animal and
bent forward (pi. 10). There is an obvious relationship be-
tween this painting and the accompanying text, which states
that the horn is on the forehead, crooked, and near the eye.
This representation is quite different from that of the Kevor-
kian manuscript (pi. 13, upper right) which provides the ani-
mal with a long, straight horn in spite of the curvature men-
tioned in the text and which is at least slightly indicated in
the Sarre manuscript (pi. 13, lower). The elephantine karka-
dann featured in the Beatty al-QazwIni manuscript (pi. 14,
upper) is show the feature referred to as the
the only one to
branch or protuberance which, following the more explicit text
of al-Damiri, is said to be on the tip of the horn. The little
excrescences on the side are probably the knots of the horn

of which al-Qazwini and al-Mustawfi speak.
Only in the early sixteenth century did artists become aware
of the fact that the horn was actually placed on the nose of
39 Ibn al-Faqih writing ca.

903 (290) (165, p. 90, footnote 1). Al-BirunI and
al-Nuwairi mention the two-horned African species and al-Dimashql that of
Sumatra and Java (85, p. 210).
40 Al-Gharnati "zeigt sich fiir das Wunderbare fast ausschliesslich empfang-
lich" (190, p. 71; see also 152, p. 68).
41 This report is repeated by al-Damiri without naming the source. Hub-
back (128, p. 3) have seen two heads with what were almost
states: "I
third horns." He refers to the Asiatic two-horned or Sumatran rhinoceros
occurring in British Malaya, Borneo, Sumatra, Burma, Siam, and Indo-China.
There are also records of three-horned black rhinoceroses in Africa (25, p. 302,
footnote; 269, p. 468; 243, p. 423).
the karkadann. However, even after this knowledge had been

gained, the animal, as such, was still often misunderstood.
Thus a miniature in the al-Qazwini of the Princeton Uni-
versity Library, giving a scene on an island in the China Sea
where the karkadann shows an animal which
is said to live,
looks more like a calf although it has a horn on its nose
(pi. 15, left). The same manuscript shows a different pic-
ture of the karkadann in the chapter dealing with the animal

(pi. 15, right). Here a curved white horn is correctly placed
on the nose of the bovine beast which retains, however, the

traditional long, straight horn on the forehead. It even shows
the spiral grooves which indicate that it is the tusk of the
narwhal. It is known that such teeth reached the Near East

and as we shall see later on that they had a certain influ-
ence on the representations of the rhinoceros horn. It seems
therefore unnecessary to assume that the Persian painter imi-
tated a picture of a Western unicorn. Had he done so, the
karkadann would also have had an equine form.
The Princeton miniature also illustrates the one and only
romantic feature of the karkadann. If one looks more care-
fully at the miniature, one notices that the head of the animal
is neither unconcernedly lowered nor defiantly raised; it rather
expresses rapt and intent listening. According to al-Qazwini,

the karkadann loves the ring dove (fakhita ^ls) and stands
under the tree in which the bird is nesting to enjoy its cooing.
The reclining animal is therefore shown listening to the sounds
42
of the small birds in the tree. Al-Qazwini states, too, that
42 This feature is obviously a romantic interpretation of a real observation.
It is based on the existence of the winged friends of the rhinoceros which
feed on its parasites and, at the same time, warn it of approaching danger.
This phenomenon has been mentioned in descriptions of the African species
in 53, vol. 12, p. 617; 162, pp. 88-89; and 263, p. 240. It is also well illustrated
by a group of square-lipped rhinoceroses from the Belgian Congo with accom-
panying rhino-birds, a variety of ox-pecker, on display in the U.S. National
Museum, Washington D. C. Dr. Herbert Friedmann, curator of the division
of birds of this museum, has been kind enough to give me an opinion about
this relationship of the two animals, especially in Asia: There are a num-
ber of kinds of birds that frequently perch on the backs of rhinoceroses,

buffaloes,and other large animals, partly to feed on the insect parasites on
the animals and partly on the insects scared up by the animals while grazing.
when karkadann does not

the bird alights on the horn, the
move its head, so as to avoid frightening it away. So far, no
miniatures have been found which show this more intimate re-
lationship of rhinoceros and dove.
In none of the texts examined was it possible to find any
reference to wings. This feature is to be found mostly on
the Spanish ivory boxes and then later in the thirteenth-
and fourteenth-century representations of the animal form-
ing a decoration on objects or buildings and in a few Shdh-
namah miniatures. With the exception of the illustration in
the thirteenth-century British Museum manuscript of Na't al-
ii ay azv an and the decorative title page of the al-
(pi. 7)
Qazwini manuscript in the Metropolitan Museum (pi. 17)
the scientific manuscripts disregard it. 43 It is, however, easy
to explain the discrepancy between texts and many represen-
tations. The texts show that the karkadann was thought to
be an exceptionally strong, fierce, and rare animal with quali-
ties which placed it in the monster class. It was not only an
exotic animal, but as al-Jahiz indicated, thought by many to

be an imaginary one. It is therefore natural that artistic im-
agination should give it the same feature which was custo-
marily attributed to representations of sphinxes and griffins. 44
Furthermore, in the period when the winged representations
of the karkadann appear in large numbers in the ornamenta-
tion of the decorative arts, even zoological books speak of
quadrupeds and human beings with wings. 45 There is thus a
The common bird of this type in Asia is the cattle egret, a rather small white
heron. This association is of mutual benefit to the bird and the large animal.
The bird gets its food in this way, and the animal is warned of approaching
danger by the action of the bird.
43 Thus the Arabic in the Bibliotheque Nationale, Arabe
Mandft al-haya*udn
2782, written in 1300 (700), shows an animal without wings as an illustration
for its chapter on the karkadann, and the same can be said about the Persian
Manafi'-i hayavan in the Morgan Library, M.500, which dates from the end
of the thirteenth century.
44 For sphinxes see 29, pp. 117-122, pi. 55. This study is quoted with reserva-
tion since its theory needs critical comment. For griffons see the Hispano-
Moresque ivory boxes referred to in footnote 4 of page 4.
45 Al-Qazwini tells us, e.g., that there is a variety of cat with batlike wings
on the Island of Java (217, vol. 1, p. 107; 218, p. 219). He also quotes a
natural tendency on the part of the decorative artist to add

wings to animals evento such well-known ones as lions and
hares 46
so as to add some fabulous and fantastic quality.
Wings are therefore not a surprising feature in nonscientific
representations of the karkadann.
story of al-Biruni according to which a governor of Asbijab sent a fox with

feathered wings to the Samanid Nuh b. Mansur (217, vol. 1, p. 451; 216,
p. 40). In his discussion of strange beings, al-Qazwini speaks of three different
island peoples having human forms and wings (217, vol. 1, pp. 448-450;
216, pp. 30, 32, and 36). These flying men are illustrated in certain al-Qazwini
manuscripts as, for instance, in the one formerly in the Sarre collection where
a green-skinned, black-bearded man is pictured with large red wings and a
white man with blue wings.
46
For the winged lion see the glass vessel in the Freer Gallery, No. 33.13
(see above, p. 3) and the Stora bucket (117, vol. 6, pi. 1291B). A thirteenth-
century gold buckle in the Berlin Museum shows a winged rabbit (226, col.
67; 167, p. 18, No. 193X).
THE KARKADANN ATTACKING OTHER ANIMALS
The two stone sculptures from Konya (pi. 3, middle and
lower) no longer treat the karkadann by itself, and they be-
long, therefore, to another iconographic group. Professor
Sarre already realized that these reliefs were not unique speci-
mens and could point to a thirteenth-century Persian luster
tile example of this group (pi. 18, upper). Further
as another
examination of medieval art objects indicates that the motif
of the karkadann attacking another animal was fairly popular
in the thirteenth and fourteenth centuries.
An obvious place to look for further examples is the ani-

mal friezes used on objects of the thirteenth and fourteenth
centuries, especially on metalwork. A winged, zoologically
undefinable karkadann running behind an elephant thus occurs
twice on the large inlaid platter of the first half of the thir-
teenth century from Syria found in Kashghar and now in the
Hermitage Museum (pi. 18, middle). 1 The same combina-
tion of animals is to be found in the upper animal frieze of
the large Syrian basin known as the "Baptistere de Saint
Louis" of ca. 1300 (700) in the Musee du Louvre (pi. 19,
upper).'- Here
even possible to establish certain facts
it is
about the "unicorn"; it belongs to the equine group and

carries on itsforehead a long straight horn with horizontal
3
striations. Further examples, this time from Iran and in the
1
The whole plate is illustrated in 233, vol. 2, pi. 153A.
-The whole basin is illustrated in 186, vol. 1, pi. 22; and 117, vol. 6, pi. 1339.
I owe the photograph of this detail to the kindness of D. S. Rice.
3 Other metal objects of the thirteenth and fourteenth centuries in whose
decorative scheme a karkadann pursues an elephant are: (a) basin of the
late thirteenth century (181, pi. 20A ; and 264, pi. 10, fig. 2 [the elephant is
unfortunately not very distinct in the illustrations]); (b) tray with the name
of the Rasulid Sultan al-Malik al-Muzaffar Yusuf (died 1295/694), Cairo,
Musee Arabe (67, pi. 47) ;
(c) brazier made for the Rasulid Sultan al-Malik
al-Muzaffar Yusuf, Metropolitan Museum
(84, fig. 1). Here two such scenes
are represented in the animal frieze. This brazier and the tray for the same
sultan of Yemen (quoted as "b") were most probably made in Egypt; (d) tray,
late thirteenth century, formerly collection of M. Edmond Guerin (185, pi. 19) ;
26
ceramic medium, are provided by two nearly identical luster

tiles in the Boston Museum
of Fine Arts and the Metropoli-
tan Museum. Both showwinged equine karkadann (pi. 19,a
lower). 4 These examples indicate that the iconographic com-
bination of unicorn and elephant is not uncommon and it can
be assumed that many more specimens will be found. There
is, however, one variant which will probably remain rare if
not unique. This is shown on a carpet painted in one of the

miniatures of the Maqdnidt manuscript, dated 1337 (738)
in the Bodleian Library (pi. 18, lower). 5 In this instance the
elephant is attacked by another gray animal of nearly equal
size, but with different ears, a peculiar forehead, and without
a trunk. It has two horns on its nose, and it uses the longer
one in front as the weapon to pierce the rump of the elephant.
There seems hardly any doubt karkadann that in this case the
is meant to be the two-horned African rhinoceros. The artist's
choice of this species for his carpet can easily be understood

when one realizes that this manuscript was executed in Mame-
luke Egypt and that the text of the nearly contemporary
Egyptian writer al-Nuwairi (died 1332) proves that people
in fourteenth-century Egypt knew about the two-horned
6
rhinoceros.
A different type of hunted animal is represented in the low-
est register of the inlaid Mosul candlesticks from the middle
of the thirteenth century in the Metropolitan Museum. It
shows in two places not an elephantine monster, but a deer

as the animal pursued by a karkadann, one resembling a horse,
(e) tray, Egypt, thirteenth or fourteenth century, Metropolitan Museum

No. 91. 1. 604, unpublished. In the outer frieze a winged equine unicorn pursues
the elephant, while the inner animal frieze shows a more bearlike unicorn as
the attacker; (f) candlestick, Egypt, fourteenth century (214, vol. 3, pi. 169);
(g) plate, Syria, fourteenth century, Eustache de Lorey collection (79, pp. 56-57,
No. 193, illustrated).
4 The Boston piece is here illustrated. The still unpublished Metropolitan
Museum tile has the number 40.181. 10.
5
The whole miniature is illustrated in 29, pi. 12c; and 125, fig. 9. Holter
established the Mameluke origin of the paintings in this manuscript (126, p. 29,
No. 77; 125, pp. 1-14).
6
201, vol. 9, p. 315.
and the other more likean antelope. 7 Such a deviation from

the more common iconography occurs also among the en-
ameled glasses of the fourteenth century, witness the covered
vessel ofMr. Kevorkian, in New York (pi. 20). On its main
band a winged onagerlike unicorn attacks with its long horn a
winged creature which is part antelope and part rabbit. This
scene is a unit in a decorative frieze that includes other fan-
tastic animals such as griffins and sphinxes. This iconographic
type occurs also in a carpet design as late as the seventeenth
century as indicated by a Mughal rug in the Widener collec-
tion of the National Gallery in Washington. Here it is, how-
ever, a realistically rendered rhinoceros that pursues the deer
8
(pi. 21 ). These are typical examples of many similar cases.
Sometimes an artist combines the two basic iconographic
types of this group and represents in a single animal frieze sev-
eral unicorns which attack both elephants and other animals.
Such a piece is, for instance, the large tray made for the Rasulid
Sultan al-Malik al-Mu'ayyad Hizabr al-dunya wa'l-din Da'ud
( 1 296-132 1/696-72 1 ) which displays in its framing frieze
three unicorns which attack an elephant, a lion, and a feline
9
beast.
might be assumed that the motif of the attacking karka-
It
dann is just another variation of the typical Seljuk and post-

Seljuk design of animals chasing each other. However this
may be in the case of other animals, the attacking karkadann
illustrates a feature that is not merely a chance creation of
7 Unfortunately the scenes are not on the side of the candlestick illustrated
in 81, fig. 87.
8 Illustrated in might be suggested that this carpet
color in 80, pi. It
36.
also shows a Along the lower edge a winged
karkadann-elephant fight.
quadruped is seen pursuing an elephant. The monster is so close to certain

representations of the karkadann that one can assume it to be identical with
this animal although it is shown without a horn. The lack of such a vital dis-
tinguishing feature can be explained by the late date of the carpet, its manufac-
ture at a court where had little appeal, and the hodgepodge character
folklore
of the piece. The animal which we have tentatively identified as a karkadann
occurs again, right behind the realistically rendered rhinoceros, so that we
would have here two manifestations of the same animal.
9
84, fig. 2, which is unfortunately not very distinct. The piece is of Egyptian
workmanship.
the artist and thus without significance, but actually the pic-
torial expression of a characteristic trait associated with the
karkadann.
The background of this particular myth is the proverbial
fierceness and strength of the animal of which even the earli-
est writers speak.
10
No wonder one reads that no animal can
11
withstand it. They all fear it and flee from it, especially since
12
the karkadann is quickly angered and never fails in attack.
"When the karkadann is in a certain territory, no other ani-
mal will graze anywhere in that territory unless there is
between it and the karkadann a distance of 100 parasangs
(about 375 miles) in all directions, because they are afraid
of the karkadann, acknowledge its and flee from
superiority,
13 14
it." It kills the lion, but the most convincing proof of the
animal's strength and fierceness is the fact that it is the arch-
foe of the giant among animals, i.e., the elephant. This un-
happy creature tries to escape it but nevertheless falls victim
to its horn. Al-Jahiz gives the following version: "They be-
lieve that the karkadann often gores the elephant and lifts it
15
upon its single horn," although he does not quite believe
this tale because, to him, it sounds more like idle talk. Never-
theless, thisaccount occurs again and again and is varied only
in details.Thus, al-Qazwini tells us that "when the karkadann
sees an elephant, he approaches it from behind, strikes its
belly with its horn, stands on its hind legs and lifts the ele-
phant until it is impaled on its horn." Hence Firdawsi applies
16
to the karg the epithet "elephant-vanquisher."
Practically all representations show the chase of a deer,
10 Al-Jahiz states that these two qualities are proverbial with the people (145,
vol. 7, p. 42), and the Akhbdr al-Sin iva 'l-Hind claims that no animal equals
it in strength, an opinion repeated by al-Mas'udi (184, vol. 1, p. 385).

11
217, vol. 1, p. 402.
12
217, vol. 1, p. 402; also al-Mustawfl.
13 This opinion also held by al-Qazwini and al-Damiri.
145, vol. 7, p. 40. is
14 referring to Ibn al-Faqih. Firdawsi also speaks of

165, p. 94, footnote 2,
the jjjl karg-i s/iir-aivzhan, "the lion-quelling rhinoceros" (102,

vol. 4, p. 318, line 478; see also p. 494, line 1605).
15
145, vol. 7, p. 42.
10
j&Jlo pil-gir (102, vol. 5, p. 202, line 1213).
antelope, or elephant with the karkadann right behind its
victim. The painter of the 'Ajaib al-makhluqat manuscript

in the Walters Art Gallery, however, illustrated the final phase
of the fierce fight when the hapless elephant was impaled on
the horn of the karkadann (pi. 22).
Al-Qazwini gives also a specific account of what happens
to the karkadann once it has vanquished the elephant; ". . .
when the karkadann is caught, it wants to free itself from the

elephant but is unable to do so. Then it falls on the ground
and both the rhinoceros and the elephant die." 17 Al-Mustawfi
uses a good reason, though based on a wrong premise, why
the horn of the karkadann cannot be withdrawn from the
elephant's body after its bulky corpse has been lifted up on
the horn: the "protuberance," the curve of which runs counter
to that of the horn, naturally makes it impossible to extract the
deadly weapon.
In contrast to the authors so far quoted, according to whom
the interlocking bodies of the two giant animals accounted
for the death of the karkadann, other writers had different
conceptions of what would happen once the elephant was
impaled. Thus karkadann
al-Jahiz says mysteriously that the
feels the presence of thedead elephant on its horn only when
18
it decomposes. Later authors have taken up this sugges-
tion and like al-Dimashqi and al-Mustawfl elaborated on it.
The last-named author, for instance, says that "the elephant's
fat gets into the eyes of the karkadann and blinds it and both
die." The second voyage of Sindbad the sailor in the Arabian
Nights also follows this version; however, after having grue-
somely told that the karkadann becomes blind from the fat
of the elephant which has melted in the heat of the sun, the
story continues, introducing a third animal: ". . . then the
karkadann lies down upon the shore and the rukh comes and
17 This story is only slightly varied by the Mandfi'-i hayavdn which has the
karkadann standing on its hind legs, raising up its front legs when it strikes
at the shoulder of the elephant. In this account, too, the karkadann cannot
pull out its horn and remains stuck to its giant prey until both perish
(5, fol. 15A).

18
145, vol. 7, p. 42.

carries it off in its talons to itsyoung ones and feeds them

with it and with that which is upon its horn." 19
A third theory explaining the death of the giant animals
is given by al-Nuwairi for whom neither the decomposed sub-
stance of the elephant nor the inability of the rhinoceros to
extract its horn from the victim's belly are apparently valid
enough reasons to account for the death of both animals. He
states that the horn of the karkadann is poisonous and thus
kills the elephant, and since the blood of the elephant is like-
wise poison, the rhinoceros, too, is doomed.""

To some
these accounts of the elephant and the karkadann
further comment should be added. First, the time has now
come to return once more to the representation of the sharav
which the artist had substituted for the karkadann in the Ber-
lin al-Qazwini manuscript (pp. 19 f. and pi. 13, upper left).
A further examination of al-Biruni's account shows that
21
the sharav, too, "attacks the elephant and cleaves it in two."
But it is not only this feature common to both Indian animals
which lead to their being exchanged for each other. Al-Birunl
concludes his report with the following statement: "Accord-
ing to popular tales it sometimes rams an animal with its horn,
raises it or part of it toward its back so that it comes to ride
between its 'upper feet.' There it becomes a putrid mass of
worms which work their way into the back of the animal. In
consequence, it finally perishes." Since these details of the
untimely death of the sharav through the decaying animal
impaled on its back are more specific than those given in the
earliest sources for the karkadann, they may possibly be the

prototype for it and may have been transferred from the
one animal to the other. In any case, it seems not surprising
that this sharing of certain features and assimilation of others
19 Littmann assumes that the stories of Sindbad the sailor were probably put
together in Baghdad in the eleventh or twelfth century. The main source
is said to be the accounts of the Wonders of India by the Persian sea captain
Buzurg b. Shahriyar of the first half of the tenth century (21, vol. 6,
PP. 747-74^)
20 201, vol. 316.
7, p.
21
45> PP- 22 an d 2 5! 44> v0 '- T > P- 2 3'

32 Freer Gallery of Art Occasional Papers Vol. I
led, eventually, to the substitution of the sharav for the

karkadann in the Berlin manuscript.
Our second comment aimed at the account of the Arabian
is
Nights which brings together karkadann, elephant, and the

rukh. Other texts dealing with the 'anqa' or simurgh, the equiv-
22
alents of the rukh, throw further light on the relationship
of these animals. According to al-Qazwini the 'anqa' lives on
some islands in the ocean on which many animals live, amongst
them, the karkadann. The 'anqa' hunts only the elephant,
large fish, and giant snakes, and leaves the other animals be-
cause they have submitted to it. 23 This preference of the
mythical bird for the elephant has some influence on the
iconography of Persian paintings. 24 It explains, for instance,
the title page of a fifteenth-century al-Qazwini manuscript in
the Metropolitan Museum. Below the roundel with the title
of the book is a little scene which shows how a winged karka-
dann warily watches an 'anqa' that is attacking its own prey,
the elephant (pi. 17). Since, according to al-Qazwini, the
karkadann is one of the animals which became subject to the
'anqa', it has to desist from attacking its usual prey. It is obvi-
ously this mental association of the fabulous bird with the
elephant and the karkadann which accounts for the combina-
tion of the three animals in the Mu'nis al-ahrdr miniature of
1341 (741) in the Cleveland Museum (pi. 8) and most prob-
ably also for the design on the carpet in the Maqamat of 1337
25
(738) in the Bodleian Library (pi. 18, lower).
22 See, for instance, the verse of the Shdh-ndmah about the simurgh stating
that "with its claws it beareth off the elephant at sight" (102, vol. 4, pp. 508-
509, line 1783; 103, vol. 5, p. 132).
23
217, vol. 1, p. 420. The Matidfi'-i hayavdn makes not only the elephant,
buffalo, and ox the victim of the simurgh, but also the karkadann (5, fol. 55b).
24 The copy of an alleged Persian drawing of the bird carrying two ele-
phants in its talons and one in its beak in 22, vol. 3, p. 90, and republished in
210, vol. 2, p. 415, is rather naturalistic and therefore of fairly recent date,
or at least a recent adaption of an older version. In its present appearance it
has hardly any Persian or Muslim character.

25 The bird on this carpet looks somewhat like an ostrich. Since the 'anqa'
is said to be the largest of the birds, it is not surprising to find that it is
reproduced in the shape of the largest living bird. This interpretation of an

Egyptian painter if we are right in assuming that he represented an 'anqa'
Karkadann and rukh are, as we have seen, the most power-

ful land animal and the most powerful bird to hunt the ele-
phant. In fact, their relationship is so close that they were

often confused, one for the other. This ambiguity of identifi-
cation is revealed in different sources and extends even to the

names themselves, as shown by the entries for kargadan in
the dictionaries of Johnson and Steingass which give the fol-

lowing translation for the Persian word: "The rhinoceros;
20
a huge kind of bird which pursues an elephant ten years old."
Unfortunately these two authorities do not quote their sources
and it is, therefore, not possible to investigate the basis for
this information. There is, however, a text which demon-
strates that the two animals were, indeed, identified with each
other. Al-Jahiz tells us that he disagrees with people who
identify the karkadann with the 'anqa' :
". . . they believe
that the karkadann and the 'anqa' mughrib are alike, although
they are used to seeing the designs of the 'anqa' pictured on
the carpets of the kings." What al-Jahiz apparently had 27
in mind was that the rugs showed a bird and not a quadruped
with a single horn on its forehead and the two animals were
therefore quite different. The erroneous identification was
natural, since the 'anqa' was admittedly an imaginary animal
and the karkadann, too, was, as al-Jahiz tells us, sometimes
thought to belong to the same category. 28 The confusion of
the two terms, probably nurtured by the fact that both animals
in question hunt the elephant, seems to have continued long
differs from an account in the History of Ahmad b. 'Abd Allah b. Ahmad

al-Farghani (died 1007 [398]), as told by Ibn Khallikan and then quoted by
al-Damiri. According to this source the Fatimid Caliph al-'Aziz had in his
collection of strange animals an 'anqa' from Upper Egypt, which "was of the
length of a heron but bigger than it in body. It had a beard, and there was
a hood on its head; there were several colours and points of resemblance to
many birds in it" (77, vol. 2, p. 196, lines 2-6; 78, vol. 2, part 1, p. 405). This
report shows at least that in Egyptian eyes the 'anqa' resembled certain unnamed
but well-known birds, a fact which opened many possibilities for the painter.
26 Vullers gives only the
149, p. 1005; 250, p. 1024; (268, vol. 2, p. 820)
meaning rhinoceros.
-7
145, vol. 7, p. 29. This, by the way, is perhaps the earliest Muslim ref-
erence to an animal carpet.
28
145, vol. 7, p. 29.
v
after the period of al-Jahiz.This might explain the fact that

al-Mustawfl calls the sharav (which is related to and thus
29
substituted for the karkadann) by the name of rukh.
It was apparently known that the land monster and the
giant bird had a common name, but instead of karkadann

(kargadan), the usual Persian term for the terrestrial ani-
mal, al-Mustawfi erroneously selected the name of the bird
rukh as the designation for both creatures. Verbal confusion
just as well as popular imagination has thus affected animal
30
lore.
29 He (191, pp. 43-44, translation pp. 30-31) uses the bird's name although
he gives al-Biruni and 'Awfi as his source, both of whom
have sharav.
150
A verbal association centering around the word karg may have influenced
the painter of the Cleveland miniature (pi. 8) in his choice of a fourth animal
for his series. It is a vulture, Persian ^j^^ kargas, literally "fowl eater";
the first syllable of this word might have formed

the mental connection with
karg-karkadann wiiich together with the elephant and the simurgh form the
content of the miniature.
THE HUNT OF THE KARKADANN
The third iconographic type is represented by scenes in
which the unicorn is itself attacked by man.
If one takes the
Arab authors happening seems well-
at their face value such a
nigh impossible in view of the ferociousness of the karkadann.
Al-Damiri is quite specific in this respect and states that "it
is very hostile to man; when it him or hears his voice
smells
it pursues him, and after having reached him kills him with-
out eating anything of the corpse." This motif is not as fre-
quent in the accounts of the rhinoceros as, for instance, those

of the death struggle with the elephant, but it is nevertheless
found in several important texts. However, according to the
character of the hunter's tale on which they are based the
stories vary a great deal. Ibn Fadian, for instance, quotes the
account of the Volga Bulgars of how the rhinoceros uses its
horn to lift a rider from his horse and to throw him repeatedly
into the air always catching him again with its horn until the
wretched victim is killed. On the other hand this writer as-
sures us that the beast never attacks the horse of the rider. 1
According to al-Nuwairi an equal enmity to man is shown by
the two-horned African variety of the animal, although the
mode of attack is said to be different. Once the hunter has
been noticed by the rhinoceros he is forced to rescue himself
by a quick flight up a tree, a strategem which in many cases
does not help, since the monster breaks the tree by assault and
kills the man. Only when the man urinates on the ear of the
animal is he able to put his attacker to flight. On the other

hand, the horn is not always regarded as the murderous
weapon of the animal, for Ahmad Tusi reports that the hunt
2
of the rhinoceros is difficult because its voice is deadly.
1
134, Arabic p. 34, translation p. 76.
2
134, p. 78 (footnote 1 of the preceding page). Actually the squealing or
quacking noise made by a rhinoceros when he human
scent and is alarmed
gets
has been compared to something between the bark of a dog and the quack
of a duck (128, p. 11); the same observer speaks of the "squeaking noise"
35
Fanciful as some of these accounts are they nevertheless

demonstrate the great awe in which the animal was held and
the difficulties encountered in hunting it. As a matter of fact,
these statements were based on experience and observation
and they are therefore fundamentally true. Yet in spite of all
hazards
and imaginary man successfully attacked the
real
rhinoceros and killed it; however, the task was so formidable
that it formed a theme in legends and later on was even
thought to be worthy of historical record. Both these types
of reports have been glorified by painters.
The scenes in which a legendary hero kills the "unicorn"
are mainly represented in the Shdh-ndmah illustrations. Al-
are not too common,
3
though such pictures several of them
have been discovered in the course of this investigation. There
must be many more which have not yet been published or are
not even recorded.
The first hero in the Shdh-ndmah to perform this deed of
prowess is Gushtasp while exiled in the land of Rum. When
Main, a Ruman noble, asks for the hand of the second daugh-
ter of Caesar, he is told that he would be accepted only after
having killed the monster. Being unequal to performing the
task, Mirin enlists the help of Gushtasp who, on foot, van-
quishes the beast on his behalf without revealing the secret. 4
Gushtasp's son Isfandiyar kills two such animals as the first
of the seven feats he has to perform on the way to the Iron
Fortress where his sisters are kept prisoners. 5 When Iskandar
is in the land of Habash (Ethiopia) he encounters a whole
pack of the African variety of the animal and slays their
of the undisturbed animal (128, p. 12), and he mentions also the "tremendous
snorts, more like an engine blowing steam than anything else" (128, pp. 6
off
and 11). In both cases he is referring to the Asiatic two-horned rhinoceros

observed Malaya, the behavior of which is not different, however,
in British
from that of the Indian and African species. Blanford (48, p. 474) speaks of
"a peculiar grunt" as the only sound produced by the great Indian rhinoceros.
3 None is illustrated in the study devoted to the miniatures of twenty Shdh-
ndmah manuscripts in the Leningrad libraries (108, pis. 1-50).

4 The story is told in
102, vol. 4, pp. 304 ff. and the killing of the animal :
on pp. 311 ff . ; 103, vol. 4, pp. 533 ff. and 337 ff. respectively.
5
102, vol. 4, pp. 490 ff. ;
103, vol. 5, pp. 119 ff.
No. 3 I. The Unicorn Ettinghausen J7
6
leader. The fourth hero to combat the animal successfully
is the mighty hunter Bahram Gur, who performs this feat
while he is disguised as the ambassador of the Shah of Iran
to the court of Shangul of India. 7 How popular belief evalu-
ated the immensity of the task of killing a rhinoceros is shown
by the fact that it is put on the same level as the killing of a
dragon which in the case of all four Shdh-ndmah heroes fol-
lows as a subsequent exploit.
When one looks over the miniatures illustrating these fights
with the monster, the differences in the representations of the
same animal again become obvious. They include some of the
new and unexpected
types discussed on earlier pages, but also a
species, a wolflike beast, which is characteristic of the Shdh-
ndmah. The cause of this profusion of types cannot be blamed
on the unbridled imagination of the artists who might, quite
erroneously, be suspected of having represented the animals
according to their fancy. It is the double meaning of the basic
term and the verbal ambiguities of the description which lead
to uncertainty and thus to mental confusion. While the poet
could be vague and suggestive, the painter had to find a solu-
tion which was often ingenious, though the zoological recon-
struction from literary tidbits is usually at great variance with
nature, and, of course, also with representations in other
manuscripts.
In Firdawsi's text the animal is called <_5" jf which, owing
to the lack of diacritical marks in early manuscripts, when ap-
plied to a quadruped can be read gurg (wolf) or karg
as
(rhinoceros). The first, being a common word, suggests itself
more readily. To make
the issue even more confusing the
text nowhere precise enough to enable the reader to choose
is
between the two. In the stories of Gushtasp and Isfandiyar

the word is rhymed with suturg (large) and buzurg (big),
thus pointing to gurg (wolf), and in the accounts of Iskandar
and Bahram Gur with targ (helmet), tagarg (hailstone), and
marg (death) indicating the form karg. 8 In the poem, Gusht-
6
102, vol. 5, p. 200; 103, vol. 6, p. 149.
7
102, vol. 6, pp. 36 ff . ; 103, vol. 7, pp. 121 ff.
8 Wolff (280, p. 704) points to these different rhyming words in his discus-
4
.
asp himself is made to express his astonishment that foolish

men should have given the name of gurg to a monster which
he regards as a dragon. 9
No wonder that the illustrators of
the Shdh-ndmah have mistaken the wolf for the rhinoceros. 10
The verbal ambiguity is cleared up, however, when the char-
acter of the animal is analyzed and, in all four cases, is found
to be of the same general type : It is a huge dark-colored
monster, like an elephant, of ferocious nature; and it charges
its foes furiously so that all animals and even brave men
avoid or flee it. Characteristically, elephants are no exception
to the rule. It terrorizes the whole countryside and victory
over it is a matter of royal acclaim and honor. What makes
it definite that the animal is the rhinoceros even when its
Persian name is, on account of the rhyme, to be read gurg

is the fact that it has but one horn, sometimes described as
being dark or even black. 11 The only misleading conception
12
of the poet is to compare this horn with that of the stag. It
is this poetical expression that induces the artist in some cases
to depict the animal in antelopelike fashion. But even the
ambiguities which pointed either to a wolf or a stag were not
the only ones to harass the painter since Firdawsi speaks also
sion of
karg.
<S ^ B, which, according to him, should for various reasons be read
11
102, vol. 4, p. 308, line 362, and p. 310, line 386.
10 The same applies also to translations like those of Mohl (who always
speaks of "loup") and the Warners; in the case of the animal killed by Bahram
Gur, Messrs. Warner realize, however, that the word "ought perhaps be read
as karg, rhinoceros" on account of the rhyme (103, vol. 7, p. 121, footnote 1).
Noldeke definitely states that the animal killed by Bahram Gur is a karg,
rhinoceros, not a gurg, wolf (200, p. 47).
In view of the ambiguity of the Persian text, the rendering of ^ as
"loup" or "wolf" in the most widely used translations and finally the wolflike
representations of the animal in Persian miniatures, it is not surprising to find
that Western captions of Shdh-ndmah illustrations nearly always speak of a
"wolf."
11 Another proof that the animal was understood as rhinoceros is the caption
of the Demotte Shdh-ndmah miniature (pi. 9) which does not use the ambiguous
term ^ but j-^" ^ '
12 See 18, footnote 29.

p.
of the claws of a lion and two tusks like those of an elephant

13
or boar.
When an artist was faced with the dilemma of illustrating
such a strange beast, one way out of the difficulty was to fol-
low the text as literally as possible and combine all the main
features in portraying the animal. This happened in the
Isfandiyar scene in the manuscript of 1429-1430 (833) in the
Gulistan Palace in Teheran (pi. 23). 14 Here the artist shows
two large wolves with long tusks, leonine paws, and a single
antlerlike horn on the head of the male animal. Whatever
one might think of these two creatures from a zoological
point of view, which indeed does not matter here, as an artist
and illustrator the painter accomplishes his task with imagina-
tion and spirit.
The gurg or lupine type in Shdh-ndmah illustrations can be
traced back as far as the fourteenth century in two paintings
showing the exploit of Bahram Gur. One is from the Demotte
manuscript in the collection of the late Mrs. John D. Rocke-
15
feller and the other from a small-sized Mongol manuscript
16
now in the 25 ) The choice of this animal-
Freer Gallery (pi. ,
type for scenes with this particular king is revealing, because

the text calls the beast killed by him not a gurg, but a karg.
Quite obviously both terms mean the same beast.
In some instances, the descriptive text cannot be found to
justify certain details in an illustration; however, such refer-
ences will appear elsewhere in the epic in accounts of other
heroes battling with the rhinoceros. Thus Firdawsi does not
speak of the tusks of a boar which the artist of the Rocke-
13 These tusks Jlj^ j1->o are such a vital feature that Gushtasp takes
them with him after he has slain the animal (102, vol. 4, p. 312, line 407;
103, vol. 4, p. 338).
14 About the manuscript see 43, pp. 69-71, pis. 43-50; 1851-
154, vol. 3, pp.
1852, vol. 5, pis. 869-874. The miniature (first published in 43, p. 71, No. 49,
and illustrated on pi. 47B), the two paintings in the Demotte Shdh-ndmah, and
one of the miniatures in the Myers collection (pi. 24, lower), are
to my knowl-
edge the only Shdh-ndmah scenes showing the fight with the lS* ^ so far
published.
|r,
54, p. in, No. 53 and fig. 26; 194, fig. 10.
16 No. 30.10A, unpublished.
4o Freer Gallery of Art Occasional Papers Vol. i
feller miniature depicts; they are mentioned only in the case

of the gurg killed by Gushtasp. 17 All this indicates that the
animals encountered by the four heroes be they gurg or
karg were understood to be the same animal and their re-
spective features were thus interchangeable to the artist. In
other respects, however, the painters of the two miniatures
took the text more literally. In the Rockefeller miniature the
awe-inspiring carcass with its very long, slightly curved horn,
is indeed of "bulk so vast" that the picture makes one believe
that "lions fled the woods before it." The painter of the Freer
miniature, on the other hand, is interested only in a simple
and direct presentation of facts and not in the emotional aspect
of the scene, and his "wolf," distinguished by size and wings,
does not arouse the hideous horror of the other painting.
In other instances the artist ignores the lupine aspect of
the beast and presents it in the shape of an antelope with a
huge single horn on its forehead. The earliest example so far
found is in the Shdh-ndmah of the first half of the fourteenth
century in the collection of George Hewitt Myers, Washing-
ton, D. C. (pi. 24, lower). Here Iskandar is shown killing
the Ethiopian karg in the form of a stag whose monstrous
character is underlined by flames above the front legs. Another
fine example is a miniature from a manuscript of the second
half of the fifteenth century, in the Bibliotheque Nationale,

Paris, in which Gushtasp is shown killing the animal (pi. 26).
Here again text and painting do not seem to match, since the
text speaks of a gurg while the painting shows an animal which
is certainly not a wolf. By now we realize, however, that this
discrepancy is not a real issue.
The problem was often solved within a manuscript by hav-

ing illustrations of two battle scenes, each with a different type
of animal, and thus the confusion in nomenclature was over-
come. In the small-sized Shdh-ndmah in the Myers collec-
tion, Gushtasp is depicted as the slayer of a lion-unicorn while
Iskandar kills a stag (pi. 24). Both animals have a large
17 The animals killed by Isfandiyar are said to have elephant's tusks (102,
vol. 4, p. 494, line 1606).
No. 3 I. The Unicorn Ettinghausen 4]
single forehead and, following a convention of

horn on their
certain Chinese legendary monsters, they also display large
flames which emanate from above their front legs. At about
the same time the Demotte Shdh-ndmah has Iskandar fighting
a leonine monster (pi. 9) while Bah ram Gur kills a wolf;
here the only feature common to both is the long single horn
on However, even this all-important trait is
their foreheads.
not always shown. For example, Isfandiyar is battling four
ordinary
that is to say hornless
wolves (instead of the two
mentioned in the text) in a Shdh-ndmah written in 1497 (902)
in the Kevorkian collection (pi. 28) while Bahram Gur kills
a single-horned antelope in another miniature of the same
manuscript 27). When one compares the animal in this
(pi.
painting with that of the Rockefeller miniature which also

shows Bahram Gur two types of karg make it
in his fight, the
obvious at once that there no iconographic consistency that
is
matches each hero with a specific animal. Other miniatures

give further proof of this. In the Shdh-ndmah dated 1605-
1608 (1014-1016) in the Metropolitan Museum, Isfandiyar
is seen killing two deer with the horns of a rhinoceros on their
18
noses (pi. 29), although in the Kevorkian miniature the
same hero is fighting four wolves (pi. 28). However, a Shdh-
ndmah of 1544 (95 1 ) in the Princeton University Library has
at least this in common with the above-mentioned manuscript
in the Metropolitan Museum, that in both cases we have an
ordinary wolf slain by Gushtasp while the antelope with the
rhinoceros horn killed by Isfandiyar in the Metropolitan Mu-
seum manuscript corresponds to a real rhinoceros in the Prince-
19
ton codex 30). All one can say from the relatively few
(pi.
examples so far tracked down is that when two combat scenes
are shown in one manuscript one represents a wolf and the
other a one-horned animal, usually an antelope. Since the
Shdh-ndmah speaks
text of the definitely of a gurg in the fights
of Gushtasp and Isfandiyar, a certain preference for the wolf-
18 The miniatures showing a fight against the karg, No. 55

*93> PP- 20-27.
(Gushtasp) on fol. 316a and No. 58 (Isfandiyar) on fol. 344a, are not illustrated
in the catalog.
19
106, p. 1, No. 1 (56 G).
\
like beastmight possibly exist in the miniatures showing these

two heroes. After the early sixteenth century the one-horned
animal reveals features of the rhinoceros, be it only in the
correct position of the horn or in the whole body structure.
How the different animal types are distributed among the
manuscripts so far described demonstrated by chart A. 20
is
The comparative frequency of the various karg scenes can

best be judged by taking a large collection of Shdh-ndmah
manuscripts as a basis of investigation. Fortunately we have
the descriptive catalog of the manuscripts in the Leningrad
21
libraries available for such a survey. These libraries own 21
illuminated manuscripts dating from 1333 to the nineteenth
century and have in a single manuscript as few as seven and
as many as 192 miniatures. One can therefore assume that
they provide a fairly representative cross section of the Shdh-
ndmah iconography. Of these 21 manuscripts three showed
a single combat scene with karg per volume and four have
a
two scenes. The distribution of the subject becomes clear
from chart B.
Combining the data provided by both chart A and chart B,
one can state that no single manuscript seems to have more
than two scenes. One can further assume that scenes with
Gushtasp and Isfandiyar are probably more common than
those showing Bahram Gur. The miniatures with Iskandar
are rare. Finally one can say that the more miniatures a manu-
script has the greater the likelihood is that one or two of the
four scenes are represented, since the scenes were found in
manuscripts with a miniature range of 29-192 only one manu- ;
script in this group had less than 50 miniatures. On the other

hand, manuscripts in which the scene is missing contain only
from 7 to 64 miniatures and only three of these contain
over 50.
20 This chart is only approximately correct, since not all paintings in the
now cut-up manuscripts are known; I have also not been able to examine the
manuscript in the Bibliotheque Nationale which may contain another karg
miniature.
21
109, pp. 1-122. I am indebted to Dr. Sidney Glazer of the Library of
Congress for having kindly translated for me the pertinent passages of the
Russian text.
;
The fight with the monster, our third iconographic type,

is not restricted to illustrations of the Shah-namah. It is also
to be found in the decoration of an implement of daily use
from outside Persia, namely, a richly decorated unglazed
water jar from Mosul. In two slightly varying scenes the
hero kills a winged unicorn of the equine type, though with
the front legs of a carnivore, the hind legs of a bird of prey,
22
and a tail ending in a dragon's head (pi. 31, left). It is hard
Chart A.- -The distribution of animal types in eight Shah-namah manuscripts
<V O +-> 1 O +-1

Lib:
Boo
orkian
iceton
Q 5 O
o o ~<2
rS
s- ,13 "55 rt o\ iv.
c,/^ hd -t-> w +j *
" "
. 2 2 2 ? :f
Heroes G o ft
Gushtasp lion antelope wolf wolf

Isfandiyar .... wolf wolf rhinoceros antelope
Iskandar lion? stag
Bahram Gur.. wolf wolf antelope
Chart B. 77/<? frequency and distribution of karg miniatures in Leningrad

Shah-namah manuscripts
00 ^ rt
^.2
doo S3
O vo S3
.00" ^ ^'e
tjoc 00
Heroes
Gushtasp . . . X
Isfandiyar . . X
Iskandar . . .
Bahram Gur X
to say which legendary person

is respresented. Three of the
23
heroes described in Shah-namah are purely Persian; the
only
the fourth, Iskandar, the Dhu 'l-Qarnain of the Koran, would
be known in an Arab country. But there is another, better
approach to the problem. Professor Sarre compared the motif
on the jar with the Persepolis relief showing a king fighting
a "unicorn" (pi. 31, right). He was right in assuming a re-
vival of an ancient motif. However, instead of regarding a
22
229, columns 8-n; 231, p. 228, pis. 105-106. In one scene the killer seizes
the horn of the animal ; in the other, one of its front legs.
23 Gushtasp, Isfandiyar, and Bahram Gur.
piece of architectural sculpture from outside Mesopotamia as

the possible prototype, might be better to look for an object
it
from Mesopotamia itself, preferably an object of the minor

arts. Scenes with a genius or king fighting a winged equine
beast with a single horn on its forehead, as found on seals,
give a good idea of what might have served as an indigenous
model (fig. i ) 24 When one assumes such a survival of an
.
age-old iconographic theme, one should also consider a local

interpretation for the scene. One is led to this conclusion by
the description of Persepolis in Ibn al-Balkhi's twelfth-century
account of Fars. When discussing the same series of reliefs
Fig. i. Seal impression. Divine hero killing a winged bull. Mesopotamia,

ca. 1200 B.C. Formerly collection of Prof. Ernst Herzfeld. (After 123, pt. 2,
fig- 135a-)
to which Professor Sarre had been referring in his discussion

of the Mosul jar, the medieval Persian writer did not identify
the hunter in the sculpture with one of the four heroes of the
Shdh-ndmah, familiar though they must have been to him; he
called him Jamshid, because the ruins of the palaces were
locally connected with this mythical Persian king.
25
On the
other hand, Ibn al-Balkhi's identification of the Persepolis ani-
mal as a karkadann makes it fairly certain that the monster
24 123, teil fig. 135A. Compare especially the winged equine body, the
2,
seizing of the beast's forelegs which are raised, and the short garment of the
slayer. Cf. also figures 148, 151, 216, 218, and 223 for partial aspects of the
iconographic type. See also 270, p. 201, No. 580; 272, vol. 2, Nos. 323, 343, and
348; 267, pp. 232-233, No. 1 and fig. 1.
25
132, p. 127 ; 131, p. 27.
.
on the Mosul jar was given the same name (see also pp. 67 f.)
The killing of rhinoceroses in historical times is recorded
26
at a fairly early date. In 922 (310) the Bulgars told Ibn
Fadlan how bowmen armed with poisoned arrows climb up a
high tree above the resting place of the animal and kill it from
27
this safe hide-out. About the same time a Muslim ambas-
sador to the Chinese court gives the emperor a similar account
28
of the way in which the rhinoceros is hunted. As we shall see
later on, the horn of the rhinoceros was a valuable commodity
exported by Muslim traders to the Far East, and it seems
reasonable to suppose that many of these were taken from
hunted animals.
Reports of rhinoceros hunts have come down to us from at
29
least the late fourteenth century on. According to the Zafar-
ndmahy Timur once killed several animals with sword and
spear on the frontier of Kashmir and the scene is therefore
2(5
One of these reports sounds somewhat fantastic, but its theme is of literary
significance. Marvazi tells us that "one way of hunting the rhinoceros is for
a man to shelter himself behind a huge tree which the animal cannot uproot
or break, then to shout at it and rouse it; it will charge at the tree, striking
it with the horn which sticks in the tree. The animal is then unable to extract
the horn and people kept in readiness come out and kill it" (6, fol. 136a; 7,
fol. 90a). This ruse is also employed in Grimms' Fairy Tales in the story of
"Das tapfere Schneiderlein" (114, vol. 1, p. 115) in which the Brothers Grimm
followed the earlier version of Martin Montanus (died after 1566) (112, p. 144).
Shakespeare, too, states that "unicorns may be betrayed with trees" (Julius
Caesar, act II, scene 1). In the letter of Prester John it is the lion that uses
this method to kill the unicorn (242, pp. 241-242), and it is also referred to in
Spenser's Faerie Queene Although it seems rather unlikely that a
(64, p. 426).
rhinoceros was deliberately hunted in this manner, incidents of such a nature
have occurred. To quote one example: "A rhino once charged a tree up which
a man had climed to escape the beast's onslaught. The rhino's horn buried itself
eight inches deep into the trunk and split the tree four inches up and down
of the point of impact. The horn was so deeply imbedded that before the rhino
could tug itself free it was shot by a companion of the man it had charged"
(161, p. 80).
27
134, Arabic p. 34, translation p. 76.
28 This confirmed by modern writers with regard to the hunting
71, p. 118. is
of the animal in the interior of Africa along the river Batha (53, vol. 12, p. 618).
29 The Mubdrak-Shdhi (quoted in 282, p. 762) states: "In the month
Ta'rlkh-i
of Zi-l-Ka'da of the same year (ca. 1387) he (Prince Muhammad Khan) went
to the mountains of Sirmor (west of the Jumna) and spent two months in
hunting the rhinoceros and the elk."
occasionally represented in manuscripts of the text such as in

the copy of 1523 (929) in the British Museum. 30 It was quite
natural that the Mughal emperors went rhinoceros hunting.
Babur relates in his memoirs how he went to a karg-khanah
or rhinoceros haunt on the Sawati in 15 19 and set fire to the
brush to drive out the animals. which had been badly A calf
scorched by the fire and found half dead on the ground was
finally killed. The emperor concludes his account by stating
that "everyone took a share of the spoil," apparently as a
trophy of the chase. 31 This scene with the burning jungle in
the background and the cut-up, blood-stained carcass of the
rhinoceros calf in the center is from
illustrated in a miniature
a now dispersed Babur -nam ah of the Akbar period in the Wal-
32
ters Art Gallery (pi. 3 2). There seems to be an even
earlier echo of the kind of rhinoceros hunt which Babur de-
scribed, because one reads in the Mandfi': ". . . the karka-
dann resorts to reedy places where it is fierce, but timorous
and not daring when it is out; it fears the flame of fire."
Another fine miniature of a Mughal emperor in pursuit of a
rhinoceros is in the O. Sohn-Rethel collection in Diisseldorf
33
(pi. 33). It shows a hunting elephant tackling a rhinoceros
with its trunk, while Jahangir, on the second elephant, gets
his gun ready for the kill. Although the two rhinoceroses in
this miniature are not accurately drawn there is no legendary
content left in this or in any other Mughal miniature, nor is
there any doubt about the identity of the beast as had been
the case in most of the Shdh-ndmah illustrations.
30 For the text see 282, 762; for the miniature, pi. 29B another
p. 115, ;
miniature showing the same subject is to be found in the Zafar-ndmah of 1533

(939) in the India Office Library (43, pi. 91 A).
31
282, p. 762; 32, vol. 1, p. 378.
32 No. 10.596, fol. 21b, unpublished.
33 Kiihnel connects this miniature with a passage in the emperor's memoirs
according to which he killed the rhinoceros while riding on an elephant. He hit
it with a single bullet near the ear (156, p. 419). Jahangir states that "it has
often happened that powerful men
. . good shots with the bow have shot
.

twenty or thirty arrows at them and not killed" (144, vol. 2, p. 270). Kiihnel is
right in pointing out that the hunted animal is probably not a wolf (gurg) as
Rogers and Beveridge, the English translators of the memoirs say, but a
rhinoceros (karg).
THE TAMED KARKADANN
Our fourth iconographic type consists of the karkadann in
conjunction with other animals and always in a subdued state.
The earliest picture of this type known to me is from a double
frontispiece of the Mongol period formerly in the Yildiz
Library of Istanbul. 1 On the right page is an enthroned ruler
surrounded by his courtiers. The opposite page shows in the
upper register a triumphal or tributory procession with vari-
ous animals, among them an elephant followed by a pacified,
winged karkadann; the lower register shows the royal horse
with a groom, falconers, and other royal servants. While the
identity of the ruler on these two detached pages cannot be
established with the data so far published," other miniatures
do not present difficulties in this respect.
One large group of paintings shows the enthroned Sulaiman,
the biblical Solomon, surrounded by the animals and demons
under his command; sometimes he is in the company of Bilqis,
the Queen of Sheba, or she is seen on her trip to him sur-
rounded by many of the king's animals. While the frontis-
piece of a Nizami manuscript of 15 13 (919) includes as a
traveling companion of Bilqis a karkadann which looks like
3
a hybrid of giraffe and deer (pi. winged equine karka-
34), a
dann is to be seen in a court scene of Sulaiman and Bilqis sur-
rounded by animals on the double frontispiece of a ShdJi-
ndmah dated 1497- 1504 (902-910), formerly in the Schulz
4
collection (pi. 35). In the sixteenth-century examples of
1
233, vol. 1, pi. 8.
2 It could be either the king for whom the manuscript was executed, or the
one for whom the text was originally composed.
3
In the possession of H. Kevorkian, New York. The manuscript was written
by Murshid al-din Muhammad in Shiraz. The right part of the frontispiece
showing Sulaiman is not preserved.
4
239, vol. 2, The left portion is now in the collection of Dr. E. Kahler,
pi. 62.
Princton, N. J. The miniature shows the elephant not only associated with the
karkadann but also with its other great enemy, the snake (cf. 5, fol. 12b: " . . .
his, i.e., the elephant's, bitter enmity is for the snake"; see also 191, p. 33,
47
+8 Freer Gallery of Art Occasional Papers Vol. i
the "Enthroned Sulaiman" the rhinoceros is usually repre-

sented as a bovine animal with a small curved horn placed
approximately on its nose. 5 The tamed,
in the corect position
even domesticated nature of the animal is exemplified in a
miniature of the Vever collection in which the rhinoceros is
shown not only standing calmly with all the other creatures,
but with strings of bells around itsneck and a saddlecloth on
its back, 6 and is thus placed in a category with the tame ele-
phant. In keeping with the general character of the Indian
art of the period, some Mughal miniatures of Sulaiman with
his hosts present the rhinoceros in a more realistic fashion. 7
Noah was another Biblical figure who, in the ark, came in
contact with the rhinoceros. Since in Persian miniatures the
ark is usually rendered in a summary fashion, no painting has
so far come to light which shows the animal. A pair of peace-
ful rhinoceroses is, however, to be found in the rear of the
lowest deck of an ark in a Mughal miniature of the late six-
teenth century which was recently acquired by the Freer

Gallery of Art. 8
A fourth variety of the iconographic type is represented by
miniatures illustrating the story of Nizami's Iskandar-ndmah
in which Plato hypnotizes animals with the help of a wonder
organ. 9 So far no Persian miniature of this subject has come
to my attention, but there is at least one fine Mughal painting
translation p. 23. Elephant-devouring serpents on Java are mentioned in Ibn

Khordadhbeh (98, vol. 1, p. 26). For classical and later Western representa-
tion of the battle between the elephant and the serpent see 119, pp. 159-160.)
5
49. Pi- 75 ; 180, vol. 1, pi. 15.
6
180, vol. 2, pi. 90. A
found on the bovine karkadann
saddlecloth is also
lifting the elephant on its horn which
Walters Art Gallery ms. W.593 is the
(pi. 22), and in the Freer al-Qazwini manuscript (pi. 14, lower). This is
perhaps due to the influence of the iconographic type of the tamed karkadann
popular in the sixteenth century; another possible explanation would be that
the saddlecloth was put on rhinoceroses used in India for staged fights with
elephants (see below, p. 88). This second explanation seems, however, not
very likely.
7
Cf. a miniature of the Akbar period in the possession of N. M. Heeramaneck,
New York.
8 No. 48.8. The miniature is reproduced, 73, p. 24.
9 78-80.
33, PP-
of the Akbar period in which the rhinoceros, like all the other
represented animals, is hypnotized by the soothing strains of
the organ played by the Greek philosopher. 10
In all these miniatures the karkadann is tamed through the
influence of an overpowering personality and no longer shows
any animosity toward the surrounding animals. However,
even in this iconographic type the traditions associated with
the usual karkadann are so strong that it is sometimes shown
in close proximity to the elephant.
With these examples in mind every student of Persian
iconography is naturally inclined to think of miniatures show-
ing the love-crazed poet Majnun surrounded by antelopes,
deer, rabbits, lions, and tigers. Here strict personal authority
is of course not responsible for the subduing of ferocious
beasts
compassion which leads tame and wild animals
it is
alike to console a forlorn and unhappy man. The subject was

popular for Nizami illustrations in Persia as well as in seven-
teenth- and eighteenth-century India, and since the Indian artist
delighted in giving a large and varied number of animals, it
is reasonable to inquire whether or not the rhinoceros might-
be found in the peaceful company around the poet. So far

only two examples of this type have turned up. The first, in
the possession of Nasli Heeramaneck, of New York, is prob-

ably from the lateAkbar period (about 1600) and shows a
Majnun
vast assembly of animals which in this case surround
and Laila who has come to see her lover at last. Among the
host of visitors one notices two rhinoceroses which, though
not acurately rendered, are easily identified (pi. 36). The
other miniature, now in the collection of Cherif Sabry Pasha,
seems to be from the eighteenth century. Here the scene is
much reduced and only a lion and a rhinoceros are portrayed
with the famous lover. 11 worth while noting that this, a
It is
fairly recent example, coming as it does from the country of

origin of the rhinoceros, still shows the animal in the tradi-
10 182, vol. pi. 181. The painter of this Nizami manuscript in the Dyson
2,
Perrins collection is Madhu Khanahzad.

11
224, pp. 151-152, No. 116 and pi. 49.
:
tional bovine shape, although with a horn now properly placed

on its nose.
There are finally two further groups of representations of
the rhinoceros which are variants of the type so far described.
In these pictures the karkadann is surrounded by other ani-
mals which it leaves in peace though no superior power is
present to keep it in check. One small group is peculiarly

Indian and shows "the Animal Kingdom" or at least some
outstanding examples of the native Indian fauna. The Freer
Gallery of Art owns a delicately drawn miniature of the early
seventeenth century, 12 which presents in a skillful composition
a large array of animals among which two fairly realistic
rhinoceroses are to be found (pi. 37). Less ambitious in
scope and artistic intention is a miniature of about 1700 in

the IslamicDepartment of the Berlin Museum where a much
smaller number of quadrupeds, including a rhinoceros, is
1
arranged in a decorative manner.
A combination of animal heads which, at times, includes
that of a single-horned creature, is the main feature of the
"talking tree" or of the tree on the Island of Waqwaq. Thus
a miniature of the Waqwaq Island from an al-Qazwini manu-
script in the Boston Museum
of Fine Arts shows, on a tree,
amongst the strange "fruits" which make the waqwaq sound,
the head of the karkadann; in spite of its bovine shape, it is
characterized as a rhinoceros by the curved horn on its nose
(pi. 38).
14
A miniature in a late fifteenth-century Shdh-ndmah
in the Bodleian Library illustrates the "talking tree" which
tells Iskandar of his approaching death; in this case the head
of the unicorn is shaped like a dragon and shows, in its Far

Eastern aspect, no relationship to any of the standard types
"'
known in the Muslim world. 1
This legendary tree is usually thought to be reflected in
12 No. 45.29. Size: 233x119 mm., unpublished.

13 pi.
39, 19.
14
52, pp. 22-23, No. 24a, pi. 13a.
15
29, pi. 38. A similar animal head is found on the tree of the Island of
Waqwaq in an al-Qazwini manuscript of 1545 (952) of the A. Chester Beatty
collection (29, pi. 37a).
No. 3 I. The Unicorn Ettinghausen 5 1
conventionalized scroll or tree designs with animal heads, such

as are to be found as decorations on a number of medieval
and postmedieval objects. " Without wanting to enter here
1
into a discussion as to whether these compositions are actually

connected with the "talking tree," it still seems appropriate to
refer at this point to the one example so far found which
includes unicorns : the plaque from the Madrasa Muqaddamiya
in Aleppo (pi. 6). As in the case of all other designs the
animal heads are at the end of the scrolls which cover the
surface and which, in this case, are arranged in strict sym-
metry, thus creating the impression of a tree.
16 i59> PP- 26 ff. 12, pp. 69-72.

;
ADDITIONAL FEATURES OF THE KARKADANN
A fuller understanding not only of the physical form and
the various iconographic settings of the karkadann, but also
of the ideas and legends associated with it, calls for elabora-
tion of our account of the animal. After all, Marvazi said
in the early twelfth century: "People say many things about
this animal," and this makes it imperative to investigate its

lore further.
According to al-Jahiz, the karkadann belongs to a numeri-
cally small species, since the dam has but few offspring and
1
her period of gestation is as long as that of the elephant. The
karkadann was even thought to be the least abundant animal. 2
The Mandfi' substantiates this by asserting that the short-
lived dam has only one offspring during her lifetime. The
male, however, seems to have a better life expectancy at
least in the opinion of al-Qazwini who gives him a life span
of 700 years. Since, according to the same author, no weapon
has any effect on it and no animal can withstand it, violent
death comes to the male animal only as a sequel to its fights
with the elephant.

In strange contrast to the alleged rarity of the karkadann
is amount of information supplied by Muslim authors
the great
which even includes the age when it reaches sexual maturity. 3
Its scarcity is, however, a good reason for the high prices
which some of the products derived from it brought on the
market. Its most precious part is the horn. Al-Jahiz is the
1
145, vol. 7, p. 40. Al-Jahiz, as recorded by al-Mas'udi (184, vol. 1, p. 387),
thinks that the period of gestation karkadann is 7 years, al-Qazwini
of the
and al-Ibshihi (99, p. 267, footnote 1) mention 3, and al-Gharnati, 4 years.
Modern zoologists speak of 17 to 18 months (53, vol. 12, p. 615; 48, p. 474).
The gestation period of the African and Indian elephant is estimated by various
authorities to last between 18 and 23 months (243, vol. 1, p. 373).
2 Marvazi quoting
al-Jahiz (6, fol. 134b; 7, fol. 88b). See also 217, vol. 1,
p. 402, and 5, fol. 14b.
3 This information is provided by al-Qazwini who sets the date after
50 years
of age.
52
first Muslim writer to refer to certain of its innate qualities

by stating that it is harder and nobler than the tusks of the
4
elephant. Since the horn of the rhinoceros is actually only
agglutinated hair, it cannot, of course, be maintained that it
is harder than ivory, but this assertion points at once to the

exaggerated attributes of this part of the body. The Akhbdr
al-Sin wal-Hind is the first of a long line of sources to inform
us that the Chinese used it for making highly valued girdles.
The horn depended on the pattern, which was
quality of the
said to appear as a light figure of a man, peacock, fish, or
other animal on a dark background and which comes out when
the horn is polished. 5 Most authors from Ibn Khordadhbeh
on agree that the figures are to be found in the inside of the
horn after it has been split. 6 Al-Mas'udi gives perhaps the
most extensive account of these girdles. According to him the
horn is usually white and shows black figures, although some-
times the designs stand out in white against a black back-
ground. 7 "With the help of leather straps girdles are made
of these horns on the model of gold and silver ornaments.
The emperors and grandees of China value this adornment
above everything else so that they pay as much as two and
4 145, vol. p. 38.
7,
5 According to Ibrahim b. Wasif-Shah (98, vol. 1, p. 160). The date of this
author is given differently. Ferrand, following Seybold (98, vol. 1, p. 137),
assumes that he wrote about A.D. 1000 and in any case before 1031 (422).
Brockelmann stated formerly (57, vol. 1, p. 335) that he lived toward the
end of the thirteenth (seventh) century, while later on he attributed his
work to a period before 1209 (606) (57, suppl. vol. 1, p. 574).
47; 98, vol. 1, p. 29. The authors who speak of figures in the split
6
136, p.
horn include Ibn Fadlan, al-Idrisi, al-Gharnati, Ibn Bukhtishu', and al-
Damiri. Passages in al-Mas'udi and in the Arabian Nights seem to reflect the
same belief. In one place al-Damiri quotes the Spaniard Abu 'Uraar ibn 'Abd
al-Barr (died in 1071 [463]) that the figures come out after the horn has been
cut off.
1, pp. 386-387. His list of figures on the horn includes the rhinoceros
7
184, vol.
itself,and he mentions also "other animals of its region." Al-Nadim's informant
told him that the most frequent designs were those of flies and of fish (192,
vol. 1, p. 349). Ahmad Tusi mentions also a lion (134, p. 77, footnote).
Al-GharnatI adds to the repertory a gazelle, different kinds of birds and trees,
besides other "wonderful things" not specifically named. Al-Damiri follows
him as usual.
5
even four thousand dinars. The clasps are of gold and the
whole is of extraordinary beauty and solidity. Sometimes one
applies different inlays of precious stones with long golden
nails." Al-Damm
adds a further detail when he says that
thin flat pieces of the horn are applied to the girdle.
The statements of the Arab authors are fully borne out by
the Chinese writers, who inform us about the use of the horn
in the official attire of the T'ang period. 8 In addition, there
are fortunately still extant several fragments of Chinese
leather girdles with applications of thin plaques of rhinoceros
horn, now preserved in the Shosoin in Nara, which are only
slightly earlier than the report in the Akhbar al-Sin wal-Hind
9
(pi. 39, left). None of these T'ang girdles has inlays of
precious stones, but a silver clasp from one of them has come
10
down to us.
In view of the large demand for the horn in China, it is
not surprising to find thatits price was as high as 3,000 to

11
4,000 mithqals or dinars of gold, and that Marvazi calls it
12
the most valuable export article to the Far East. The price
depended on the quality of the pattern and on the popularity
of the material at a given time. Thus already in the middle
8 Laufer questions the use of the horn

165, p. 142, footnote 4. in "official"
girdles of the T'ang dynasty, an opinion first held by Bushell, and states that
they were restricted to the use of princesses. Dr. Schuyler Cammann was
kind enough to check this for me. He supplied the following information:
"The T'ang History (Hsin T'ang shu) discussing the everyday dress or
informal attire of officials says: first and second rank officials' belt plaques
'
. . .
used gold, officials above the sixth rank (third to sixth) used rhinoceros horn,
officials above the ninth rank (seventh to ninth) used silver, and common people
used iron.' Meanwhile a reference in the T'ang dynastic statutes {T'ang hui
yao) under the date of 710 A.D. says that for Imperial audiences and state
banquets, first and second rank officials are permitted to wear (dress acces-
sories of) jade or t'ung (-t'ien) rhinoceros horn, while the third rank is per-
mitted to wear carved rhinoceros horn or striped rhinoceros, or jade, etc. These
references make it very clear that the rhinoceros horn was used as a precious
substance on a par with gold and jade."
9 and which our plate made. Detailed
147, vol. 1, pi. 20, vol. 7, pi. 33, after is
descriptions are given in Japanese and English.

10 After 147, vol. 1, pi. 20.
11 According to Ibn Khordadhbeh, Marvazi 135b), al-Damiri quoting
(6, fol.
Abu 'Umar 'Abd al-Barr, Ibn al-Wardi (98,
b. vol. 2, p. 412), etc.
12 183, Arabic p. 5, translation p. 17.
of the ninth century A.D. Ibn Khordadhbeh tells us that the

price varied from 300 to 4,000 dinars. Al-Nadim is the
only writer who gives an explanation for a drop in the market
price since he had the good fortune
have as informant a to
monk from Najran, who in 987 (377) had returned from a
visit to China and could give him a first-hand report about his
experiences. This man told him that the price per ounce of
horn had fallen from 5 minas of gold to 1 ounce of gold and
even less. This drop was caused by a royal decree which abol-
ished the custom of appearing with girdles of rhinoceros horn
and required instead girdles of gold and similar costly ma-
13
terial. Hence, following the law of supply and demand, the
much-decreased demand had suddenly caused a depression in
the valuable commodity.
There are some other exotic uses of the horn mentioned
by Arab authors. According to the tenth-century Hudud al-
'Alam the chiefs of the Waqwaq province of China "wear ex-
tremely precious necklaces of rhinoceros horn," 14 and Ibn
Bukhtishu' tells us that the kings of China hang it on them-
selves against evil things. The Shosoin again provides an
example form of several small containers used as pen-
in the
dant ornaments, made from rhinoceros horn one horn-

shaped and the other consisting of two square boxes. They
are combined with small amber and crystal balls (pi. 39,
15
right). The same collection also contains a small fish-shaped
13 Ferrand interpreted
192, vol. 1, p. 349, vol. 2, p. 184; 98, vol. 1, p. 130.
al-Nadim's passage as referring to tribute in the form of rhinoceros horn, but

the use of girdles with applied pieces of horn in Chinese court attire (see above,
footnote 8) indicates that the text refers apparently to the latter. This is also
the way in which the German editors of the Arabic text translated the passage
(192, vol. 2, p. 184).
In view of the prices quoted by the medieval authors it is worth while noting
that in 1935 the value of the horn on the Calcutta market was about half its
weight in gold. A single horn retrieved from the poachers fetched 150 pounds
and still higher prices have been known (241, p. 1229).
14
129, p. 84. This is corroborated by al-Qazwini (217, vol. 2, p. 30, line 15)
who states that the inhabitants of Sandabil (Kan-chou in the Kan-su Province)
adorn themselves with "elephant and rhinoceros bone." De Goeje already
thought that this might be the horn of the rhinoceros, mounted in gold and worn
as an amulet (178, p. 87).
15 58a and 58b for other horn containers like-
147, vol. 7, pi. 51; see also pi.
wise square or horn-shaped. These containers are called gosu in Japanese.

Freer Gallery of Art Occasional Papers Vol. i
pendant of rhinoceros horn. 16 Another usage for the horn

was to apply it to thrones in thin pieces. Al-Gharnati, fol-
lowed by al-Darniri, does not specifically say where such
thrones were made, but from the fact that he tells us about
this custom in connection with the Chinese girdles, the infer-
ence is that it is used in the Far East. A more commonplace
service rendered by the horn is first referred to by al-Biruni,
who tells us that it furnishes the material for the handles of
knives. mentioned by several other authors, even
This use is
in modern times, and it must have been quite popular and
widespread. Only one medieval source gives a more specific

description of a particular feature of these handles. After
having stated that the horn of one of the African species is
white but for a black streak from the tip to the center of the
base, Marvazi adds: needed in sawing it so that each
"Skill is
handle contains a black circle." Let us hope that we shall be

fortunate enough one day to find a medieval Muslim knife or
sword with a rhinoceros handle as described by the twelfth-
century Persian author.
The Akhbdr al-Sln wal-Hind is the only source to give
some information about the commercial transactions to pro-
cure the horns: ". . . they are bought in the State of
17
Dharma with cowrie shells which are the currency of this
land." Marvazi tells us of a man in Balkh who had imported
the horn of the African species. The same author mentions

also the export of the horn to China. Some come definitely
from a certain region in India, but in another place he states
that the port of export to China was Basra, which seems to
point to the African horn.
The hide of the rhinoceros, too, is commercially used. Ac-
cording to al-Damhi it is worked into coats of mail against
which weapons have no effect. Al-Mustawfi, moreover, men-
tions that leather belts which sell for 3,000 to 4,000 dinars
are made "from the handsome cuirass of leather which it (the
16 56L Modern Chinese regard the horn as a potent aphro-

147, vol. io, pi.
disiac (118, pp. 377, 386, 387 and 395).

17 Apparently Bengal and Bihar (183, pp. 147-148; 17, pp. 52-53 and map 2).
karkadann) has around its middle." Since the quoted high

price is the same as that for the celebrated Chinese belts
made with horn plaques, it may be surmised that al-Mustawfi
who alone mentions the leather variety, confused the two
18
types. Various parts of the body served also as remedies or
talismans. Thus its gall is used for fumigations to dispel evil
19 20
smells, the left eye against the shaking fever and the stings
of scorpions, 21 while the right eye is a talisman against pains,
22
jinns, and demons. The most effective part of the body is
the horn and especially its legendary protuberance. It is a

remedy against colic and labor pains, epilepsy, paralysis, and
spasmodic contractions of muscles. 23 It helps against the evil
eye, unties knots, makes hot water cold, and prevents a horse
from stumbling. 24 Al-Qazwini demonstrated its powerful
magic by telling a story according to which particles of the
horn (or rather of its "knot") mixed with dust were scattered
on a band of robbers, thereby preventing them from rising
from their camp to plunder a caravan. The same author, fol-
lowed by al-Mustawfi, mentions also that "When it is near
poisoned food or drink, it nullifies (literally, breaks the strength
of) poison." A variation of this notion is found in Ibn al-
Wardi who speaks of knife handles made of the horn which
in the presence of poisoned food become moist and agitated.
While it is thus certain that the karkadann had magic prop-
erties, no religious connotation seems to have been connected
with it. 25 The pious Muslim was only concerned whether it
18Another possible explanation for the high price of the leather belts men-
tioned by al-Mustawfi would be that the Chinese girdles were made of
rhinoceros hide. This assumption is disproved, however, by a girdle in the
Shosoin which is made of moleskin (147, vol. 1, notes to pi. 20).
19 Ibn Bukhtishu'.
20 Al-Qazwini, al-Damiri.
21 Al-Damiri.
22 Al-Qazwini, al-Damiri.
23 Al-Qazwini, al-Mustawfi, al-Damiri.
24 Al-Damiri.
25 are sometimes found in the descriptions of
Such religious connotations
animals. For instance, al-Mustawfi says of the silkworm: "This worm is a
mighty example of the manifestation of the power of the Artificer with whom
none may be compared, the Creator of 'Be, and it was,' who from the slime
was permissible to eat its meat or not. 26 The interpretation

of dreams in which the animal occurred pointed also to its
purely secular character, since according to al-Damiri the
karkadann supposed to represent a powerful, tyrannical
is
king or to indicate war and trickery. This seems like final

proof that the karkadann represents an evil force.
If one were to summarize all the main thoughts which
could have been in the minds of persons who saw the picture
of such an animal, one could state The karkadann is a very :
rare, large, and powerful animal of India and other distant

countries. It is of evil and ferocious nature and quickly an-
gered. It does not allow any other animal in its neighborhood
and pursues and kills even the mightiest of them all, the
elephant. Only kings of great magic power like Sulaiman
could subdue it, and only heroes and mighty hunters were able
to kill it. The outstanding part of its body is the horn on its
forehead with which it kills the animals. Various designs are
visible in this horn which in China is used for highly prized
ornaments. It serves likewise as a powerful remedy and talis-
man. Thus the karkadann stands for fierceness, evil physical
force, and magic power.
of such an insignificant worm produces such elegant garments ..." (191,

p. 59, translation p. 41) ; or of the peacock: "In every one of its feathers
are present red, yellow, green, gold, blue, and other colors such that the
hand of the painter cannot portray a picture of the like. And glory be to Him!
How great is His majesty, and how wide His power, and how perspicuous
His proof!" (191, p. 109, translation p. 77).
26 The author of the Akhbdr al-Sin iva'l-Hind ate the flesh and al-Mas'udi
speaks of it as permissible food since it is a kind of buffalo (184, vol. 1,
p. Al-Damiri quotes al-Zamakhshari's opinion in this respect: "The

386).
evident factor which speaks for its lawfulness is that the animal eats trees
and is a ruminant. Its being an enemy of man does not preclude it from
being eaten because the hyena treats man as an enemy and it is eaten. But
if it were proved that it is born from the horse and the elephant, it would
be forbidden. And this is remote." Modern hunters have reported that the
meat of the rhinoceros is quite palatable (258, p. 147).
r
OTHER UNICORNS IN MUSLIM LITERATURE
During the preceding discussion of the karkadann we have
at various times had occasion to refer to animals which either
became fused with the karkadann or split off from it and
turned into new species. We therefore have to say something
more about these other creatures.
1
Al-Jawhari, Ibn Bukhtishu', and al-Damin state in their
discussion of the karkadann that the Arabs call it harish (actu-

2
ally Ivrysh) or, vice versa, that the harish is the karkadann.
1 Quoted by 191, pp. 40-41, translation p. 29; 77, vol. 1, p. 285; 78, vol. 1,
p. 525 (s.v. ) ; see also 77, vol. 2, p. 327, line 9.
2
fJt^r" w i tn or without the article is the spelling in most manuscripts and
published texts. Kraus (153, vol. 2, p. 67, footnote 15), Stephenson in his
translation of al-Mustawfi (191, p. 28) and Jayakar in his translation of
al-Damin (78, vol. 1, p. 525) transliterate it as harish, and this is also the
form in 9, fol. 89b, where full vocalization is given. That the correct form of
the name was doubtful in the Middle Ages is indicated by the many variants.
The Ibn Bukhtishu' manuscript in the Bibliotheque Nationale has al-haris,
while the Morgan Mandfi' writes jurish (with a vowel for the first consonant,
but with no diacritical points for the letter ya) the two al-Qazwini manu-
;
scripts in Mr. Kevorkian, the Berlin Museum,

the Beatty collection, those of
and of the Freer Gallery (dated 1789) have j rysh which is also the spelling .
of the Wustenfeld edition of the Arabic text (217, vol. 1, p. 392). An

al-Qazwini manuscript dated 1570 (978) in the Harvard College Library and
another dated 1664 (1075) belonging to H. Kevorkian have ^ ws -
As has already been pointed out by Laufer (165, p. 124, footnote 1) harish
may be connected with the following statement by Cosmas Indicopleustes on
the rhinoceros: "The Ethiopians in their own dialect call the rhinoceros arou
or harisi aspirating the alpha of the latter word, and adding risi. By arou they
designate the beast as such and by arisi ploughing, giving him this name from
the shape about the nostrils, and also from the use to which his hide is turned."
(76, p. 358).
The Babylonian Talmud {Hullin 59b) knows of a unicorn called KHp which
is described as "the hart of the forest 'Ilai" (69). The Aramic name might
explain the enigmatic <jfep as a garbled form of C?Js which like BHp (69)
would then be the last part of the Greek words "pLvoKepcos and fxovoKepws. Ibn
Bukhtishu' (3, fol. 28A) states that the Syriac-speaking people call the karka-
dann-harish L>j j rima, which reproduces the Syriac rayma corresponding to
the Hebrew re'em. The Princeton University Library manuscript does not
59
Actually they are two quite distinct animals, although both

are unicorns. The first writer to speak about the harish as a
separate animal is apparently al-Tawhidi who says: "It is
a certain small animal of the size of a kid or lamb and very

quiet, but it has such strength of body and swiftness of mo-
tion as to baffle a hunter. It has in the middle of its head a
solid and straight horn with which it strikes all animals.
Nothing can subdue it. It is necessary to use a stratagem for
seizing it, namely to expose to its view a young virgin or a
young girl. When it sees her, it jumps into her arms as
though intending to suck her milk, which is a natural mark
of affection ingrained in its nature. When it jumps into her
arms it sucks her breast, though there is no milk in them, with
such a gusto that it is overpowered by intoxication like the
intoxication from wine. While it is in that state the hunter
comes and ties it up firmly with a rope being itself motionless
on account of this dodge." 3 The Mandfi' 4 overcomes the
difficulty inherent in the preceding account in which the harish
is sucking the milkless breasts of a virgin. It introduces in-
stead of the virgin a beautiful girl from a brothel ( J&~*)

who tempts the animal with her breasts. The animal sucks
them for about an hour and then falls asleep from the milk.
have rima but basrima. F. Rosenthal kindly suggested to me that the ba is
probably a mistaken preposition taken over from some Syriac context, while
the sin is just a misinterpretation of an overlong space between the ba and the
ra in the Arabic script.
The Arabic translation of the Physiologus refers to the name of the animal
which has all the characteristic features of the harish as lo daiya (l^J JUj
As F. Rosenthal suggests, this may be merely a corruption of Ujj.
3 The passage is quoted by al-Damiri (77, vol. 1, p. 285; 78, vol. 1, p. 525).
F. Rosenthal informs me that it is taken from al-Tawhidi (255, vol. 1, p. 184)
as was also recognized by Kraus (153, vol. 2 p. 67, footnote 15). Ahmad
Tusi likewise tells the story of the hunt with the help of a virgin. The animal
in this case is, however, the karkadann (134, p. 77, footnote 1). Zeki Validi
Togan translates J-ij^ ^
weiblichen Geschlechts," but, judging from the parallel
in the account of the strategem as "Saugling
texts, it probably means
"a virgin girl." Ahmad Tusi gives a variant of the account by stating that
when the rhinoceros smells the girl, it faints and so do all the other karkadann
in the steppe.
4 fol. 54a.
5,
.
The story ends with the exclamation "Allah knows best,"

thus expressing a certain incredulity about the story. It is
probably for this reason that al-Qazwini leaves out the story
of the temptress; he contributes, however, to the lore of the
animal when he states that it is found in the swamps of
5
Sistan and the land of the Bulgars and he finishes up with
giving the medicinal properties of its body, thus exchanging
6
the folkloristic tale for a scientific myth. The story of the
girl and the harish is given by Ibn BukhtismV 7 and also by
'Awfi (as quoted by al-Mustawfi) who calls the animal
( 8
qat a{\*^& ) 'Awfi's passage and also the Manafi' dissent from
.
the usual belief in the single horn by stating that the animal
has two horns on its forehead. 9 It is for this reason that we
see a small kidlike animal with two horns sucking the breasts
of a young woman in the miniature illustrating the chapter of
the harish in the Morgan Manafi' manuscript (pi. 40, lower)
No other representation of the peculiar hunt of the harlsh
has so far turned up. The usual iconographic type of the
animal is the isolated figure of a goat- or antelope-like creature
with one horn on its forehead such as those found in the
and sixteenth-century al-Qazwini manuscripts in the
fifteenth-
Kevorkian and Beatty collections (pi. 41, middle right and
lower). A more monsterlike harish is included in the Berlin
5
217, vol. 1, p. 392. Jacob (141, pp. 166-167) identifies the fleet-footed
animal with the saiga antelope of southern Russia the adult specimens of which
cannot be overtaken by horses or greyhounds.
6 "When a man suffering from quinsy drinks its blood with hot water, his
obstruction will meat cooked with the centaury plant, when
come up at once. Its
eaten by a man suffering from colic, will cure him at once. When his ankle-
bone is burnt and its ashes with its fat are placed on the bleeding(?) artery,
the pain stops ..." This medical lore is repeated by al-Mustawfi (191, p. 41,
translation p. 29) and al-Damiri.
7 Under the heading
8 I I
9 >PP- 48-49, translation pp. 33-34. The vocalization of the animal's name
is uncertain.
9 In the table of contents of 'Awfi's Janvdmi* al-Hikaydt, by Muhammad
Nizamu' d-din (198, p. 257), a goatlike animal is mentioned which is hunted
with the help of a girl whose breast it sucks. It is obviously the same animal,
but in contradiction to the information given by al-Mustawfi the text says that
it has a single horn.
manuscript (pi. 41, middle left), where it is shown walking

on its hind legs ; this follows the al-Qazwini text which states
that the animal does most of its running on two feet. In the
Sarre manuscript the harish resembles the painting of the
wolf (pi. 41, upper). 10 If this
is not a mere coincidence, the
choice of the bodily form may have been conditioned by the

wolf monsters in the Shah-namah. Another, more clearly de-
fined influence from a different iconographic source is to be
found in the al-Qazwini manuscript of the Princeton Uni-
versity Library. Here the miniature of two snakes, one climb-
ing up a tree and devouring young birds and the other com-
ing out of a rock, in no way follows the text about the harish
it purports to illustrate (pi. 40, upper). This can, however,
be explained by one of al-Jawhari's two identifications of the
animal which are quoted in al-Damiri's text. While the second
describes the harish as a karkadann with the claws of a lion,
the first regards it as "a certain speckled species of serpent."
The painter of the Princeton manuscript, or rather the origi-
nator of the iconographic type which he copied, must have
preferred this dissenting opinion of al-Jawhari and the illus-
tration to go with it. In any event, we have here just as

in the case of the sharav of the Berlin al-Qazwini manuscript
which claimed to be a karkadann the substitution of a differ-11
ent animal from another text for the usual iconographic type.
Another animal which becomes separated from the karka-
dann to lead an existence of its own is the sinad. As we have
pointed out before (pp. 15 f.), it shares with the karkadann
the peculiar feature that the young one puts its head out of
its mother's womb before it is born to find outside nourish-
ment (pi. 16, lower) Al-Jahiz in his account of the karkadann

.
12
is the first to report this yarn on Indian authority, while
10 Disregarding the horn, the harish and the wolf in the "Sarre al-Qazwini
manuscript" differ mainly in color, the harish being reddish brown, while the
wolf is dark gray.
11 The Turkish al-Qazwini manuscript in the Walters Art Gallery dated
1709 (1121) likewise shows the influence of another text. The harish is shown
as a two-horned goat, which indicates that the iconography goes back to such
versions as the Mdndfi'-i hayavdn or the Nuzhat al-Qulub of al-Mustawfi.
12
145, vol. 7, pp. 40-41.

al-Biruni as quoted in al-Qazwini's chapter

13
on the sinad
applies it to an unnamed Indian animal. Al-Biruni tells us,
too, that the birth takes place only after the young one is
strong enough to run away from mother because the very its
rough tongue of the dam would separate the flesh from the
bones of the calf when licked after the birth. 14 This story is
16
then repeated by al-Qazwini, al-Mustawfi, 15 and al-Damiri
in their Marvazi, however, mentions
chapters on the sinad.
karkadann after he
the licking episode also in the case of the
has expressed his doubts about a story in which he alleges
that the karkadann devours its offspring. But he states that
the licking scene is true and adds that the tongue of the
mother is sharper than a file. 17 Al-Gharnati even speaks of
18
a big thorn on the tongue. The prenatal activity of the
foetus and the peculiar character of the tongue are there-
fore associated both with the karkadann and the sinad. 19
Although al-Qazwini and his followers, al-Mustawfi and al-
Damiri, stated that the sinad is shaped like an elephant (see
p. 16 and pi. 16, lower), the knowledge that its peculiarities
are also those of the karkadann induced the illustrator of
the Princeton al-Qazwini manuscript to give to the sinad the
traditional features of the karkadann (pi. 42, center). First
of all, this particular sinad is not elephantine, but has the
bovine shape usually given to the karkadann. On the nose
13 217, vol. 1, p. 396.

14 Al-Jahiz tells only that the karkadann can run fast after its birth and is
thus able to fend for itself; there is in this connection no reference to the
danger of its being licked by its mother.
15 translation p. 18.
191, p. 26,
16 part
77, vol. 2, p. 41; 78, vol. 2, i, pp. 81-82.
"6, fol. 134b; 7, fol. 88b.
18 This account is copied by al-Damiri. Al-Gharnati (again followed by
al-Damiri) informs us also that the kings of China torture people by having
them licked by the karkadann, which separates the flesh from the bone.
19 The belief that the young rhinoceros runs away from its mother right after
its birth is just the opposite of the actual condition. The calf of the white
rhinoceros accompanies its dam until it is practically full-grown, that is, until
the birth of the next calf (243, vol. 1, p. 434; 162, p. 88). Hunters have
reported that the calf remains withits dead mother up to 2 days (25, p. 306).
Photographs of a female rhinoceros followed by her calf have been repeatedly

published (38, pp. 143 and 147; 97, p. 174).
of the dam are placed the horn of a rhinoceros as in the

Princeton karkadann miniature (pi. 15, right) and the addi-
tional long, straighthorn on the forehead (but without the
grooves of the narwhal tusk) and finally, since some authors ;
point out that the horn, teeth, and hoofs of the karkadann
grow while the embryo is in the mother's womb, 20 these parts
of the body were likewise given to the young sinad. The scene
itselfshows the mother reclining on the ground after having
given birth and the young calf running away with great speed.
A similar spectacle has so far not been found in the iconog-
raphy of the karkadann, although the literary sources would
make it theoretically possible.
21
Another type of unicorn is called jliîU shadhahvar or
22
erjT, aras which, according to al-Qazwini is to be found in
the remotest regions ofRum. 23 On its single horn are said to

be 42 hollow branches. They form a kind of Aeolian flute
because the wind produces cheerful or plaintive sounds when
20 Stated in the Mandfi'-'t hayavdn and by al-Ibshihi and al-Damiri. This

has been confirmed by contemporary zoologists as far as hoofs are concerned
(53> vol. 12, p. 616), but the newly born animal has no teeth and there is
only a smooth, rounded hard boss for the base of the horn (59, p. 433).
21 This is the spelling in the "Sarre al-Qazwini manuscript." jlâ^l-i sha-
dahvar, as given in the Wiistenfeld edition (217, vol. 1, p. 398), is also to be
found in the fifteenth-century al-Qazwini manuscript of the Berlin Museum,
and the two Beatty manuscripts. However, the al-Damiri text has j] Â^Li
(77, vol. 2, p. 48) and Jayakar transcribes it therefore as shad-hawar (78,

vol. 2, p. 97). The form jlôjL- sadahvar is to be found in the Kevorkian
manuscript (fifteenth century) and the Princeton University manuscript.
22 The name occurs first in the early tenth century in the works attributed to
Jabir Haiyan, the alchemist (153, vol. 2, p. 67). Kraus transliterates it as

b.
aras and assumes it to be derived from Greek opv oryx, which according to
b
Aristotle was single-horned (27, book 2, chapter 1, p. 499 line 19). The ,
Wiistenfeld edition of al-Qazwini has ^fj) 'rs (217, vol. 1, p. 398). The fif-
teenth-century al-Qazwini manuscript of Mr. Kevorkian has Jj> j|

'rsh. We
follow the transliteration of Kraus.
Without wanting to stress the point (since it may be a mere coincidence)
it can be pointed out that the name of the Angolan or black-faced impala ante-
lope in Bushman is a:ras (243, vol. 2, p. 557).
An early treatise on the properties of the horn of the aras attributed to
Jabir b. Haiyan is, unfortunately, lost (153, vol. 1, p. 154, No. 1994).
23 217, vol. i, p. 398.
passing through them. These are so pleasant that other ani-

mals are attracted by This yarn is first found in the writings
it.
24
attributed to Jabir b. Haiyan about 900 A.D. Al-Damiri
increases the number of ramifications on the horn from 42 to
72 and al-Mustawfi makes the animal into a fierce carnivore
which chooses its prey amongst the animals which gather
around it. 25 Al-Mustawfi may not have invented this fiendish
trait of the shadhahvar, since he could have transferred it
from the siranis which has 12 orifices in its snout and pro-
duces a flutelike sound when drawing in its breath. In this
case, too, animals gather around the monster eagerly listening
26
to its sound before they are captured and eaten. As noticed
before, the transfer of a curious feature from one fabulous
animal to another is quite common in animal lore.
The earliest illustration of this "unicorn" that has come
to my knowledge is in the "Sarre al-Qazwinl manuscript." It
represents an antelopelike animal with a tremendous horn

from which jut out 14 short branches projecting in alternate
directions (pi. 42, upper). Just as al-Damiri increases the
number of ramifications of the horn given earlier by al-
Qazwini, so does the painter who copied the "Sarre manu-
script" and produced the volume which is now in the Beatty
collection. In his miniature the horn has grown so long that
24 These writings state also that Plato was supposed to have captured the
aras, a feature not reported in the Greek sources (153, vol. 2, p. 68).
25 31. Here the text of the 'Ajd'ib al-makhluqdt
191, pp. 44-45, translation p.
is,however, interpreted to mean that the animal has 2 horns, each carrying
21 hollow branches.
20 translation p. 31, which gives the erroneous form jJl j-<-w siwanis.
191, p. 44,
The correct spelling can be inferred from the fact that the story is taken over
from al-Qazwini (217, vol. 1, pp. 397-398) where the animal is called jJj
siranis.Furthermore, G. Jacob (141, p. 167) has shown that the siranis is the
Greek siren he also followed G. Hoffmann in pointing to the connection between
;
siranis and the siranas in the Syriac Physiologus (see 207, p. 51, No. 38, foot-
note 1). The siranas is described as a sea animal with seven openings in its
mouth into which it places the seven toes of its feet when it wants to sing.
Then all the other animals gather to listen. A long beak with many holes from
which different notes emanate is also to be found in the lore of the qaqnus,
"phoenix" (191, p. 119, translation p. 86), an Indian bird whose name comes
from Greek kvkvos "swan" (F. Rosenthal).
at 17 points short branches come out in opposite directions.

The horn of the animals in the al-Qazwini manuscripts in the
collection of Mr. Kevorkian and Harvard College Library
the
show no branches, only a great many holes (pi. 42, lower
left). The Berlin copy of the same text shows an animal in
whose horn neither branches nor holes are indicated, thus
demonstrating in one more instance how in this manuscript the
iconographic exactitude has become weakened (pi. 42, lower
27
right). The first painting to show the shadhahvar sur-
rounded by intently listening animals is in the al-Qazwini
manuscript of 1545 (952) in the Beatty collection (pi. 43,
lower). In this case the antelopelike animal carries an enor-
mous horn, now curved, which has both hollow branches and
holes. The Princeton al-Qazwini. manuscript illustrates the
same scene, but gives up the traditional antelopelike appear-
ance of the shadhahvar; instead one finds a fantastic leonine
animal derived from the Chinese hsieh-chai (pi. 42, upper). 28
A fifth quadruped unicorn mentioned by Muslim writers 29
is much smaller than the others so far mentioned, being a

yellow hare with a single black horn on its head. Its name is
^1 jxj) al-mi'raj. According to al-Qazwini it was given to

Iskandar by the inhabitants of the Sea-Serpent Island (
Jazirat
al-Tinnin) in the Indian Ocean after he had killed the serpent.
In spite of its small size even the mi'raj has, apart from its
single horn, an important feature in common with the karka-

dann, namely, that every wild animal runs away from it when
it sees it. Ferociousness seems altogether the foremost feature
shared by all the Muslim unicorns. The "Sarre al-Qazwini
manuscript" contains a miniature of the animal which closely
follows the details given in the text (pi. 44, lower). In con-
trast to it the often not-too-specific Berlin manuscript shows
27 It will be recalled that it was this manuscript which substituted the sharav
for the karkadann.
28 The hsieh-chai is a white lionlike monster with a single horn. It should
be distinguished from the ch'i-lin and the pai-tse (68, p. 108 and figs.4 a-c).
29 (who the animal al-
217, vol. 1, p. 113, lines 9-11; 218, p. 230 calls
"mi'raj") ; 77, vol. 2, p. 392. The vocalization of the first syllable is uncertain.
a hybrid animal which looks more like a fierce hound than a

hare upper) 30
(pi. 44, .
The "minor unicorns" described in this chapter have so far

been found only in purely scientific texts. Since their folklor-
isticand literary interest was apparently limited, they did not
pass on into the larger stage of belles-lettres and they never
gained wider popular appeal. The unicorns in Muslim art
can therefore be assumed to represent, in practically all cases,
the various versions of the karkadann. The validity of this
last assertion will, of course, become much more obvious if
it can be shown that a medieval Muslim readily identified a

representation of a unicorn as that of a karkadann. Fortu-
nately we can adduce such proof. When in the beginning of
the twelfth century Ibn al-Balkhi gives, in his treatise on Fars,
an account of the remains of the palace of Persepolis known
as the "Throne of Jamshid" (Takht-i Jamshld), he speaks
also of the various showing a king (or
well-known reliefs
31
mythical hero) fighting wild beasts. This author describes
an encounter with one of these animals, a horned monster
combining the features of lion, scorpion, and griffin (pi. 31,
right) in the following way: "Jamshid is represented . . .
taking a karkadann by the horn while in his hand he holds

32
a hunting knife which he has plunged into its belly." Al-
30 One further unicorn, the bulan which is restricted to Turkestan has

probably no iconographic significance, since it is unlikely that it was ever
illustrated. According to the eleventh-century Mahmud al-Kashgharl this ani-
mal lives in the country of the Kipchak. At the tip of its horn is said to be a
deep hole in which snow and rain accumulate. When the male animal is
thirsty the female kneels down and lets him drink, and he does the same for
his mate (173, vol. 1, p. 346, lines 11-14; 58, p. 120).
31 In earlier investigations Sarre and Herzfeld called the animal-killer a
king (232, pp. 134 ff. 228, p. 13, ;
pis. 16 and 17). More recently Professor
Herzfeld has taken exception to this and has called the figure the ancient mythi-
cal hero Krsaspa (122, p. 257).
32 Ibn al-Balkhi speaks in his account of a lion, a
132, p. 127; 131, p. 27.
wild ass (j^") and a karkadan, all killed by Jamshid. Three such royal
animal fights occur in the Palace of Darius, the tachara, where a lion, a bull
(the "wild ass" of Ibn al-Balkhi), and a lion-scorpion-griffon with a long single
horn, obviously the karkadan, are killed by the king. (232, p. 134; 228, pi. 16.)
For various reasons these three reliefs seem to fit the description of Ibn al-Balkhi
better than the four of the Palace of Hundred Columns which includes also
though this particular animal has not the slightest resemblance

to a rhinoceros and is even totally different from the usual
representations of a karkadann as we have analyzed them in
earlierpages of this investigation, the fact that it is a quad-
ruped monster with what appears to Ibn al-Balkhi to be a
single horn on the forehead is proof enough to suggest a
karkadann. Judging from this experience one can assume that
Ibn al-Balkhi would have likewise identified other unicorns on
medieval Muslim objects or buildings as representations of
the karkadann.
At this point, when our survey of the authentic Muslim uni-
corns has come to an end, some comments are appropriate on
what one might call the "apocryphal" unicorns. There are
first those which do not seem to occur as regular illustrations
to the classic Muslim The outstanding examples of this
texts.
category are of Far Eastern inspiration. The most common
type is derived from the Chinese ch'i-lin, a composite beast
with a dragon's head, the body of a stag often covered with
scales, and
bushy tail; they are to be found in the decorative
a
arts of the Timurid and Safavid periods, particularly in the
illuminations, bookbindings, and carpets. It is not impossible
that they are sometimes meant to represent the karkadann,
especially as the iconographic setting in a frontispiece of an
al-Qazwini manuscript 17) allows such an interpretation.
(pi.
The fact remains, however, that none of the Timurid or

Safavid copies of the 'Ajd'tb al-makhluqat so far found use
the ch'i-lin as illustrations for the chapters on the karkadann
or for any of the other unicorns. Since the Persian late six-
teenth- or seventeenth-century Cosmography of the Walters
33
Art Gallery uses a ch'i-lin-like animal in the water to repre-
sent the "sea ox" or "water bull," it would appear that the
Chinese creature does not seem to have a fixed counterpart
in Islamic zoology.
Nevertheless the ch'i-lin was, together with other Chinese
a scene with a lion-bird-griffon (232, pp. 135-138; 228 pi. 17). Plate 31,
right, illustrates the relief in the south doorway of the west wall in Darius's
palace. I owe the photograph to the kindness of Dr. Erich F. Schmidt.
33 1, fol. 181b.
)
monsters, rather common in the Islamic repertory. In certain

instances one can even see why the artist chose it for a specific
setting. To quote two examples : the scene, frequently found
in sixteenth-century art, of a lion attacking a bull or stag, de-
rives from the motif of a lion killing a bull which may be
found in the staircase reliefs of Persepolis. 34 In the Achae-
menian representations the victim is given in strict profile so
that only one horn shows. This feature must have struck the
imagination of a Turkish artist specializing in black and white
line drawings. Since he employed this technique in imitation
of Chinese ink paintings and used as the left part of his com-
position a Chinese dragon (lung) it was only natural that he
substituted a Chinese unicorn for what looks like a unicorn
35
in the traditional 45 This change was
Persian design (pi. ) .
possible because this artist was mainly interested in beauty of

line and of design as such, and not in literal rendition of a
subject.
Our second example is the fight between the ch'i-lin and
the feng-h uang (phoenix), occurring inmore or less stylized
form in the art of Persian books of the sixteenth century 36
and on some Caucasian dragon rugs of the following century. 37
In trying to explain this scene one could point to the Mandfi'-i
hayavdn which states that the karkadann is one of the victims
of the giant bird simurgh. 38 This text is, however, rare and
the myth is so uncommon that it can hardly be taken as the
source for the motif of a bird descending on a unicorn. Neither
does Chinese art help us to find the proper interpretation.
There the feng-huang and ch'i-lin are not regarded as mutu-
ally antagonistic and they are only rarely placed together in
a design.Furthermore, the earliest example a dish of the
K'ang Hsi period ( 1662- 1722 89 in which I found the bird
swooping down on the fabulous quadruped is later than the
34 228, pi. 21.
35 Freer Gallery of Art, No. 48.17.
36 11.
16, fig.
37
265, vol. 2, pi. 3 ; 257, No. 4, fig. 2.
38 fol. 55b.
5,
39 The two animals appear already on the opposite sides
124, pi. 44, fig. 3.
of a blue and white vase of about 1400, but no obvious connection between
the two seems to exist (158, pi. 27).
6
earliest known Persian examples. seems therefore not very

It
likely (though not impossible) that the prototype of the Mus-
lim scene is Chinese, although the animals themselves are of
Far Eastern origin. There is, however, another explanation
possible. In the art of Iran and her bordering countries dur-
ing the fifteenth and sixteenth centuries fabulous animals of
Chinese derivation were often placed in landscapes or in a
decorative setting. The lung and the feng-huang are the most
common of these, and they are usually shown as being engaged
40
in combat. The ch'i-lin is also rather frequently found. The
easiest explanation would be to assume that the dragon was
replaced by the ch'i-lin, which to a Persian artist was just
another fierce-looking monster. There are, however, certain
settings of the Chinese animals that admit also another ex-
planation. Sometimes the ch'i-lin is placed in the space be-
tween the feng-huang darting down on the rearing dragon. 41
To certain artists copying such a design the nature of the
main motif was not known, or it did not seem significant enough
to preserve it. Thus, at times, an artist chose only a section
of the scene with the result that now the feng-huang com-
batted the ch'i-lin. Be that as it may, the new combination is
first found in the Persian art of the sixteenth century, but
since a bookbinding of about A.D. 1430 shows a feng-huang

42
battling a winged griffon, a feng-huang fighting a ch'i-lin
might also be possible in this century. There seems little doubt
that the Caucasian carpet weavers borrowed the chinoiserie
motif from a Persian model; these rugs show many such in-
spirations and the weavers were in any case more interested
in color and pattern than in realistic and meaningful scenes.
All this leads one to assume that the combination of the two
Chinese animals is accidental and the scene devoid of a
special meaning.
A second group of unicorns are actually spurious; they look
like unicorns but are really two-horned animals. Their picture
should in most cases be interpreted as showing the horns in
40 I intend to deal with this motif in another paper.

41 212, vol. pi. 1128.
6,
42
16, fig. 9 -
No. 3 I. The Unicorn Ettinghausen 7i
strict profile so that they cover each other and appear as one.
A case in question is the early fourteenth-century miniature
of a yahmur in the Freer Gallery of Art which originally be-
longed to a now scatteredhayavdn manuscript Mandfi'-i
43
(pi. 46). Here the horn ison the brow of the clearly set
head and thus looks like a single horn. Another miniature
portraying two cervine animals from the same manuscript, in
the Minneapolis Institute of Art, shows clearly that the artist
could draw two antlers or horns in proper perspective if he
wanted to do so. 44 There is nothing in the nature of the
yahmur which would oblige the artist to depict it with a single
horn. The modern dictionaries designate it as a wild ass or
45
onager, without an allusion to horn or horns. The text of the
Mandfi' itself, although it does not describe the physical aspect
of the animal, states at least that "its nature does not differ
from that of the deer âT." And finally the late thirteenth-
century Mandfi
1
manuscript in the Morgan Library portrays
the yahmur with two horns. There is therefore no reason to
count the yahmur as another unicorn. Whether the artist of
the Freer miniature regarded it either as a unicorn or was
under the influence of the unicorn iconography and thus repre-
sented it as such, or whether he only wanted to represent it
in strict profile, all these are unanswered questions and will
probably remain so. Still the case of the yahmur is significant
because it proves that an animal does not have to be a unicorn
even if it looks like one.
Another instance is a little scene on a carved Fatimid ivory
box in the Berlin Museum. 46 Here a lion is seen attacking an
antelope which is shown with only one horn on its brow. The
iconographic setting makes it impossible to regard this animal
47
as a unicorn, as was suggested in recent publications. It can-
43 No. 38.2.
44 figure on The miniature represents the "mountain ox and
245, p. 45.
cow" ^yb^ which sheds its antlers every year.

45 There perhaps a confusion with the two-horned yamur which
is j
al-Qazwini compares with the wild ox (217, vol. 1, p. 405).
40 On the left side of the cover in 155, vol. 3, pi. 254.
47
64, p. 430; 55, P- 276.
not be the mighty karkadann, which overpowers all animals;

and as to the kidlike harish, which seems to be the most likely
identification, we find that Ibn Bukhtishu' states: ". . . it is
stronger and runs faster than the lion." The ivory box shows,
therefore, a two-horned antelope in strict profile.
This poses now the question whether we have not been too
rash in regarding the noncaptioned animals as unicorns, or es-
pecially as rhinoceroses. However, our conscience can be put
at ease for two reasons. First, the scenes of a unicorn chas-
ing an elephant can refer only to a karkadann. Second, the
single-horned animals in the captioned illustrations of zoo-
logical and Shdh-ndmah manuscripts form such a large and
varied body of rhinoceroses that they easily identify or at
least provide the raison d'etre for the comparatively few un-
designated specimens on buildings and works of art. There
may be one or two examples which may eventually be proved
not to be a karkadann, but the rest will undoubtedly stand
up under further scrutiny and remain in this category.
THE LORE OF THE UNICORN IN THE MUSLIM WORLD
COMPARED WITH THAT OF OTHER CIVILIZATIONS
For the sake of clarifying the complex lore of the unicorn
the preceding pages were restricted to a study of the physical
character and iconographic setting of various one-horned ani-
mals world of Islam and to the associations which people
in the
there had formed about them. The Muslim East is not, how-
ever, the only region which has concerned itself with the uni-
corn. This animal is universal with wide ramifications through
the ages and in various cultures, most of which antedate the
1
birth and rise of Muslim civilization. With this fact in mind
one can assume that since the Islamic world has always been
ready to accept and integrate ideas which conform to its con-
cepts, an interrelationship must have existed between the be-
liefs held by Islamic writers and those of other civilizations.
This is all the more the case because the lands of the caliphate
occupied a strategic position between India and China to the
east and the lands of the classical heritage and the medieval
civilizations to the west, all of which have fostered myths
about the unicorns. By investigating a possible interchange our
understanding of the prehistory and the significance of the
unicorn pictures will obviously be deepened. In view of the
complexity of this ramified and widespread material, many
of our explanations can be nothing but hypotheses; but at least
in a few instances a new and fairly well-documented insight
into the growth of the myth seems to have been gained.
A. THE ANCIENT ORIENT

There is nothing in the Muslim
which would indicate
texts
that any particular idea about the karkadann was taken over
from one of the pre-Muslim civilizations of the Near East.
It seems, however, possible that the ancient arts might per-
haps be responsible for the equine type of karkadann.

]
On this subject see 242, pp. 26-48.
73
The Mosul jar in the Berlin Museum on which this particu-

lar type occurs for the first time belongs to an archaistic class
of pottery the ornamentation of which often shows icono-
2
graphic survivals from the Babylonian period (pi. 31, left).
As we pointed out in an earlier passage the scene represented
might very well have been inspired by an early seal. It is
irrelevant whether or not the ancient model actually represents
unicorns. Ibn al-Balkhi's interpretation of the Persepolis
sculpture shows that the medieval Persian was ready to see
a karkadann in any animal which looked as if it were repre-
sented with only one horn. Likewise, it does not matter that
the old prototype may not have been a horse at all, but a
bull of equine appearance.

3
Ibn al-Balkhi shows again how
loose the medieval reinterpretation of an old motif could be,
since in another Persepolis relief he mistook a bull for a wild
ass. The Mosul jar follows the old oriental motif rather
closely. might be surmised that later on and in other regions
It
the equine unicorn was recognized as a karkadann, especially
since it was popularly believed, as shown by the Muslim texts,
that the karkadann had some connection with a horse. It was
then detached from the traditional iconographic setting and
adapted to purely Muslim ones.
B. THE CLASSICAL WORLD

A few Muslim writers have acknowledged their acquain-
tance with classical texts dealing with one-horned animals.
4
Al-Jahiz states that Aristotle speaks of the one-horned
"Indian ass" in his Historia animalium. 5 Later on Marvazi
quotes two Greek authors for his transcription of 'pwoKepuxs and
2 The dependence of the artists of the Seljuk period on the arts of pre-Muslim
periods has been noted in the discussion of textiles (237, pp. 84-90), pottery,
and metal, but so far no systematic discussion has appeared. The influence
comes mostly from Sasanian prototypes; the unglazed pottery of Mesopotamia
follows, however, Babylonian models.
3 Herzfeld (123, 116 and
teil 1, p. fig. 29) stresses that the shape of the bull
of this period "ist mehr die eines pferdes" ; see also 270, p. 201, No. 580,
and p. 202.
4 who,
145, vol. 7, p. 40, in turn, is quoted by al-Nuwairi (201, vol. 7, p. 315).
5
27, book 2, chapter 1, p. 499b, lines 15 ff . ; 28, pp. 218-221.
:
for the meaning of this word in Arabic. 6 The same author

presents, in addition, several other excerpts or references to
classical writers, nearly all of whom he mentions by name. In
doing so he distinguishes himself from other Muslims dealing
with the subject who only rarely use Greek sources. Marvazi
refers to Ptolemy's Geography in connection with the habitat
of the animal, and he paraphrases Aristotle's theory that a
singlehorn goes with a solid hoof and that this horn is placed
in themiddle of the head. 7 In another passage he gives a gen-
eral description of the karkadann according to a certain
^yj^J*]. Here he is apparently reproducing the section on
the rhinoceros from the Liber de animalibus of Timothy of
Gaza (^Jj^J*). 8
In the manuscript of the India Office Library the transliterated term is
rendered as cfJ^S^JJ which is translated as jîJ] ji "he with a
horn on the nose." This shows that the original Greek word must have been
l
piv6nepoos. Marvazi quotes "Ali b. Zain" (according to 6, fol. 134b) or "Ali
b. Din" (according to 7, fol. 88b) as his source for the statements of a Yunani
and also a certain <jfy* The Ali b. Zain (or Din) is, as Franz Rosenthal
informs me, probably Ali b. Rabban al-Tabari. This ninth-century writer was
born in Merv in 808 (192) as the son of a Christian scholar of Syrian origin
and with Syriac as his mother tongue. Only late in his life and after having
finished his Firdaivs al-hikma in 850 (235) did he become a Muslim convert.
Marvazi's great admiration for the Greeks has already been noted by Minorsky
(183, p. 2; cf. 57, vol. 1, p. 231, suppl. vol. 1, pp. 414-415).
7
6, fols. I35a-i35b; For the view of Aristotle see 28, pp. 218-221.
7, fol. 89b.
8 For comparison we are juxtaposing the texts of Marvazi and Timothy of
Gaza
Marvazi Timothy of Gaza
"Its size isand its
that of a horse "In size the rhinoceros is about the
habitat is and the sur-
on the Nile same as the river-horse (hippopota-
roundings of bahr al-asamm (?). On mus). Coming from the ocean, he
its nose it has a single horn, like a dwells by the side of the Nile. He has
sharp sword, with which it can split on his nose a horn like a sword with
rocks by hitting Sometimes it at-
it. which he is able to bore through even
tacks an elephant with its horn and a rock, and with this he often kills
kills it. The whole of the species con- elephants. All rhinoceri are male and
sists of males; there are no females how they generate, no one knows . .
."
and no one knows how they come into (259, vol. 3, pp. 297-298, para-
existence and how they are born" graph 45).
(6, fols. I34b-i35a; 7, fols. 88b-89a).
Marvazi follows Timothy very closely, although the details of expression are
slightly different. Steier (248, columns 1339-1341) has, however, pointed out
It also happens that a reference to a specific classical writer

turns out to be incorrect. Thus Ibn Bukhtishu' thought that
the stories about the harish (which he equated with the karka-
dann) went back to Aristotle. Actually they derive from the
Physiologus.Although Ibn Bukhtishu' was wrong, his error
indicates, at least, that he felt he was dealing with a classical
story now incorporated in Islamic literature.
Two found both in al-Biruni and
classical motifs are to be
Marvazi. These authors tell us that the frontal horn of the
rhinoceros becomes erect when it charges and wants to strike
with it and that the animal sharpens its horn against rocks so
that it can cut and pierce.
The reference to a movable horn occurs in Pliny's Historia
nataralis where this feature is attributed to the yale (a mythi-
9
cal animal) and to the forest bull in Ethiopia. Cosmas
Indicopleustes attributes it, however, to the Indian rhinoce-
10
ros. This myth has been explained by the late Sir Arthur
Shipley in "the practice of some African tribes who trained
the horns of their cattle to point, one forward and the other
backwards. In the imagination of travelers these developed
intomovable horns which the beast could point forward and
backward at will." 11 As the hunting accounts of the Swede
Andersson show, the myth was still current in Southwest
Africa in the middle of the nineteenth century. 12 It probably
persists to the present time.
The horn-sharpening feature is recorded for the first time,
in the first century B.C. by Diodorus Siculus, 13 and is then
that the manuscript which served as basis of Haupt's edition is rather poor.
The which Lambros followed for his edition unfortunately
better manuscript
does not contain the chapter on the rhinoceros. The text of Timothy goes
indirectly back to Oppian's Cynegetica (see the passage in 202, pp. 102-105;
for Timothy's sources see 248, column 1340). Timothy lived at the time of
the Emperor Anastasius I (491-518).
9 book
209, vol. 3, 8, pp. 54-55.
10
76, P- 358.
11
40, p. 40. For the history of the yale in classical medieval times and its
heraldic use in the West, see 89, pp. 173-199.

12 "It is believed by many . . . that it can, at will, turn the posterior horn,
the other horn meanwhile remaining firm and erect" (25, p. 304).
13 86, vol. 2, book pp. 180-181.
3,
repeated by Pliny, 14 Aelian, 15 and Solinus. 16 In classical texts

the rhinoceros sharpens its horn prior to its fight with the ele-
phant. Al-Biruni does not speak of this preparatory step in
connection with the fight between the two animals because
he himself had witnessed in India such an encounter in which
this sharpening of the horn did not take place. He like
Marvazi
gives it only in the hearsay account of a traveler in
Africa. 17 Thus this myth, like that of the movable horn, prob-
ably reached the Muslim world from the Dark Continent.
Whether it represents the direct influence of indigenous folk-
lore or its classical version is a question that cannot be inves-
tigated in this paper. It can, however, at least be stated that
the horn-sharpening motif probably originated from the fact
that the biggest horns of the species which was known to
the classical world (the white or square-lipped rhinoceros)
are usually worn flat at the front of the tip, because the horn
drags on the ground when the animal grazes. 18 In some cases
captive animals have reduced their horns to small stumps or
the horns have altogether disappeared, due to constant whet-
19
ting. Although this seems to be the right interpretation of
the old idea, classical sources replaced the correct explanation
by another which is based on the belief (expressed by
Agatharchides, 20 Diodorus,
21
and Aelian 22 ) that the horn of
14 book
209, vol. 3, 8, pp. 52-53.
15 14, vol. 432: book chapter 44 (early third century).
1 , p. 17,
] 6
Quoted by 74, p. 16 (Solinus lived in the second half of the third century).
17 There is so far no indication that Muslim authors believed in or at least
reported another African myth, namely, that "the horns are soft and pliable
when the animal is at rest, and that they at once become hard and solid
when on the move." (25, p. 304).
18
120, p. 31; 53, vol. 12, pp. 607 and 613. The identification of the classical
rhinoceros goes back to Trouessart (53, vol. 12, p. 607).
53> vol. 12, p. 617; 97, p. 182. The notion of the classical writers
19 may
have been further strengthened by the observation how the rhinoceros uses its
horn to get the bark off trees, makes wallowing pits (128, pp. 3, 6, and 14),
digs under small trees or bushes to loosen their roots (53, vol. 12, p. 611), or
"noses" with its horn and muzzle in ant-hills (see the exhibit in the U.S.
National Museum, Washington).
20 This writer of the second century B.C. is quoted by Photius
(206, columns
55-56).
21 book
86, vol. 2, 3, pp. 180-181.
22 14, vol.
1, p. 432: book 17, chapter 44.

the rhinoceros is much like iron in strength and hardness.

Starting from this assumption, the ancients concluded that the
horn can be sharpened with a kind of natural whetstone. How-
ever this may be, in Islamic literature the motif remains a
rare feature.
The next question to be discussed is whether or not the
fairly common accounts of the fight between the rhinoceros
and the elephant found in classical literature since Agathar-
23
chides and Artemidorus (second century B.C.) are the pro-
totype for this motif in the writings of Islamic authors. The
Muslim accounts are either stereotyped or very fanciful and
thus in neither case based on actual observation. They must
be derivative and the question is only from which source
classical or otherwise.
When one compares the Islamic accounts, especially the
earlier ones, with those of the classical writers, one notices
that the two differ in certain aspects. They do not have the
specific classical motivation, the fight for watering places and
24
pastures; they do not explain why the karkadann tries to
hit the belly of the elephant, which in classical literature is
25
described as the softest part of the elephant; and, more-
over, they do not mention that if the rhinoceros is unable to
strike the elephant's paunch, the elephant in turn kills it later
on with its tusks and the sheer weight of its body. 26 For
23
For a short survey of the literary references see 247, columns 1785-1786.
A fresco with a two-horned rhinoceros following an elephant was found in a
Sidonian tomb-chamber of Ptolemaic times at Marissa (Moresheth) in Palestine
(204, p. 26, pi. 10; 220, p. 67 and pi. 5, fig. 13). Since the elephant carries
a saddlecloth and is apparently guided by a Negro, the fresco can hardly
be assumed to reproduce the proverbial fight between the two animals; only
their combination in the painting could have been suggested by the classical
references.
24
Agatharchides (second century B.C.) as quoted by Photius (206, columns
55-56) Artemidorus (about 100 B.C.) as quoted by Strabo (253, vol. 7, pp. 334-
;
335); see also Diodorus Siculus (86, vol. 2, book 3, pp. 180-181 and Aelian ;
(14, vol. 1, p. 432: book 17, chapter 44). All these passages speak of the
rhinoceros.
25 Pliny (209, vol. book Solinus (74, p. 16).
2, 8, pp. 52-53) ;
26 Agatharchides as quoted in Photius (206, columns 55-56); Diodorus

Siculus (86, vol. 2, book 3, pp. 180-181) Aelian (14, vol.
; 1, p. 432: book 17,
chapter 44) ; Artemidorus as quoted in Strabo mentions, at least, that the ele-
Islamic writers, the sole motive for the fight is the ferocity
of the karkadann which prevents any animal from grazing in
its very extensive territory. There is never an allusion in
Islamic literature to the possibility that the rhinoceros might
not be victorious. Since the time of al-Qazwini, however, they
have given more detailed descriptions of the manner in which
the karkadann attacks the elephant and of the ensuing death
of the attacker as a consequence of his victory. As has already
been pointed out, the Islamic accounts lack also the common
classical feature of the sharpening of the horn on a rock prior
to the assault of the rhinoceros. All these differences seem
to point to the fact that the Muslim stories are not directly
dependent on classical literature. If this is the case, how can
it be explained that both civilizations have the same motif
but with such divergencies?
It is natural to assume that the story of the fight between
elephant and rhinoceros came from regions where the two
animals are at home, that is, either Africa or India. No Mus-
lim references pointing to an African origin of the motif have
so far turned up. It is not mentioned in the African part of
al-Biruni's report, though it is otherwise permeated with fanci-
ful stories. The account of the Egyptian al-Nuwairi, which
contains both Indian and African lore, speaks of the attack on
the elephant as taking place in India. Furthermore, nearly all
Muslim authors speak of the single-horned karkadann when
they mention the combat; only al-Dimashqi speaks in this con-
nection of the two-horned species of Sumatra, while al-
Gharnati followed by al-Damiri refers to a three-horned
variety, which again excludes the two African species which
are bicorned. Finally, various reports about the relationship
of the African animals directly contradict the Muslim myth
of a rampant and ever-victorious rhinoceros. Thus it has been
said that "even the ponderous and quarrelsome black rhinoce-
ros (which is more aggressive than the white species) will
. .invariably make off (when elephants approach), usually
.
giving vent to his fear or ire by one of his vicious and peculiar
phant can prevent the belly-slitting attack of the rhinoceros by the use of his
trunk and tusks (253, vol. 7, pp. 334-335).
27
snorts." Another authority states that when the two animals
visit the same water hole, the black rhinoceros at all times
2S
gives way to the elephant. Big-game hunters in Kenya, Brit-
ish East Africa, insist that the rhinoceros will never remain in
the presence of an elephant, since the two are deadly foes and
the latter always emerges victorious from any conflict. 29 This
belief has recently been challenged by an American zoologist
who "on several occasions saw both animals drinking from
the same watering place, with neither paying the slightest
30
attention to the other," which, at least, indicates that the
African rhinoceros is not eager to combat the elephant.
All these data speak against an African origin of the Mus-
lim myth. This does not, however, preclude the fact that ac-
counts of deadly encounters are unknown on this continent.
Andersson reports in the book on his hunting experiences in
Southwest Africa, published in 1857, that "furious battles are
said to take place occasionally between rhinoceros and ele-
phant; and though, of course, strength in the elephant is infi-
nitely superior to the rhinoceros, the latter, on account of his
swiftness and his sudden movements, is by no means a despica-
ble antagonist. Indeed instances are known where they have
perished together. At Omanbonde we were told that a com-
bat of this kind occurred not long before our arrival. A
rhinoceros, having encountered an elephant, made a furious
dash at him, striking his long sharp horn into the belly of
his antagonist with such force as to be unable to extricate
himself, and in his fall, the elephant crushed the assailant to
death." 31
Two details of this report, viz, the attack of the
rhinoceros against the belly of the elephant and his being
crushed to death by the weight of the collapsing elephant, re-
call at once the similar account of Aelian who wrote more
27 2 5> P- 2 5> footnote i.

3
28
243, vol. 1, p. 417. There is no reference to any combat; likewise Lang
in his account of the white rhinoceros in the Belgian Congo does not list the
elephant as an enemy (162, pp. 87-88).
29
252, p. 3.
30
252, pp. 3-4.
31
25, p. 124. Ken Stott, Jr., informed me that he too found that the African
natives regard the two animals as deadly enemies.
No. 3 I. The Unicorn Ettinghausen 8i
than 600 years before the Swedish explorer. 32 This might

i,
suggest that the latter's version and perhaps also the refer-
ences of classical writers to the superior strength and power
of the elephant as the decisive element in the combat are of
African origin.
The fact that we are inclined to exclude Africa as the
source of the Muslim myth leaves us now with India as the
other alternative. Only two Indian texts dealing with the
33
motif have come to my notice. They are unfortunately late
and not specific enough to clarify the issue, but being the only
indications of a possible Indian origin of the story, they
deserve a short discussion at this point.
Babur stated in his memoirs that in certain regions it was
thought that the rhinoceros could lift an elephant on its horn,
34
a belief not shared by the emperor himself. Unfortunately
he did not give any further details and we therefore do not
know which geographical area he had in mind. The editor
of the memoirs assumed that he was referring to Tramon-
35
tana. If this localization of the story is correct, it still re-
mains undecided whether it is the usual tale found in Muslim
literature or represents a local "Tramontanan version." In
the latter case it could very well be a reverberation of tales
from the not too distant Indian jungle country where the ani-
mals are at home. There is, however, still the possibility that
the emperor may have referred to India after all. While the
exact region referred to remains uncertain, the remarks of the
Mughal emperor show at least the state of mind from which
the myth arose. In one place in his memoirs he muses: "I
have often wondered how a rhinoceros and an elephant would
behave brought face to face." 36 He leaves this question
if
open, yet in another passage he states: ". the rhinoceros . .
is more ferocious than the elephant, and cannot be made

32
14, vol. 1, p. 432: book 17, chapter 44; see also 74, p. 15.
33 As Professor Geiger has kindly informed me, there is no reference to the
rhinoceros-elephant fight in classical Indian literature.

34
32, vol. 2, p. 489.
35 The region between the Hindu Kush and the Oxus River.
30
32, vol. 2, p. 451.
37
obedient and submissive." Starting from such an attitude
it is natural that a simple-minded, less critical person would
soon come to the logical conclusion about the outcome of a
fight.Thus a native hunter, who with simple arms had been
exposed to the headlong attacks of the angered rhinoceros,
and also knew of the greater docility of the elephant, could
easily come to a belief in the bitter rhinoceros-elephant fight
which ends fatally for the latter. And even if the hunter did
not tell such a yarn, it would naturally suggest itself to people
to whom the dangers of the jungle are vividly described. The
final steps in the development of the story are further exag-
gerations and the diffusion of the accounts by traders, itinerant
craftsmen, soldiers, and sailors. Naturally, such stories would
find an eager ear in the greater safety of regions removed
from, but still in relative proximity to, the jungle, such as
Tramontana. And of course, once the notion about the
ferocity of the animal became associated with the observation
38
of its solitary life, another myth resulted, namely, that the
rhinoceros does not let any other animal graze in its territory
within a radius of 100 parsangs.

The second reference to the fight between rhinoceros and
elephant is of fairly recent date. When, after a stay of 20
years in Bengal, Capt. Thomas Williamson wrote his Oriental
Field Sports in 18 19 he repeated in his chapter on the rhinoce-
ros most of the features known to Muslim writers about this
animal including enmity for the elephant. 39 William-
its fierce
son writes that he is "compelled to rely much on the report

of those residing in situations frequented by the rhinoceros,"
and although he discredits the accounts of the natives who
professed to have been present while the rhinoceros and the
elephant were fighting, he nevertheless seems to be under
the influence of the local beliefs of such deadly encounters. 40
37
32, vol. 2, p. 490.
3S
53, vol. 21, p. 612; 128, p. 5.
89
278, pp. 163-177.
40 174-175. The scene with "a rhinoceros charging
278, pp. 164-166 and
elephants" published by Coomaraswamy (75, vol. 2, pi. 22) is after a design
of Captain Williamson drawn by Samuel Howitt (278, vol. 1, pi. 11). This
picture shows at the right a large elephant whose belly has been ripped open
by the rhinoceros in such a way that the blood-stained intestines show.
.
These two sources (in the absence of earlier and more spe-
cific accounts which, it is hoped, will eventually turn up) lead
one to assume that the Muslim belief in the elephant-rhinoce-
ros fight belongs possibly to Indian folklore. Such an assump-
tion is, to a certain degree, corroborated by the fact that the
account of the fight occurs in the "Voyages of Sindbad" which
41
are strongly permeated by Indian popular stories. Here we
find, furthermore, a clear statement of how the yarn was dis-
seminated. In the second "Voyage" it is said that ". . . the
and persons in the habit of journeying about
sailors, travellers,
in the mountains and lands have told us that this wild beast
which is named the rhinoceros lifteth the great elephant upon
its horn ." Thus it seems most likely that Muslim sailors
. .
Fig. 2. Indian seal found in Tell Asmar. Middle of third millennium B.C.
(After 104a, fig. 108.)
and travelers picked up the tale in India and brought it home

with them.
As far as this writer knows, no early Indian figural render-
ing of the elephant-rhinoceros fight seems to have come to
light. It is true an Indian seal, with a well-defined rhinoceros
behind an elephant, was found in a house of the middle of the
third millennium B.C. in Tell Asmar (the old Eshnunna) in
Iraq, where must have been brought by ancient trade ( fig. 2 )
it
The fact that there is a crocodile-like gharial above the rhinoc-
eros, however, would seem to invalidate the assumption that

this is an early representation of the elephant-rhinoceros fight;
4^ The rukh stories likewise reveal Indian influence. A third Indian motif
is found in the seventh "Voyage," where we learn of the burial place to which
the elephant repairs when it feels the approach of death (23, vol. 3, p. 177).
The Singhalese parallel was pointed out by the British zoologist Sir J. Emerson
Tennent more than 80 years ago (256, pp. 181-182).
it would be wrong to assume a relationship between the two

animals without considering the third, in this combination. An
even more forceful argument is presented by a seal impression
found in Mohenjo-daro on which is a procession of four wild
animals, the elephant following the rhinoceros, and the latter
preceded by two beasts which Sir John Marshall defines as a
tiger or leopard and another catlike animal. Compared with
the Tell Asmar seal, the position of the two animals in which
we are here interested is reversed. All we can say, therefore, is
that on seals of the Indus civilization the elephant and the

rhinoceros occur together with other animals, without any fixed
42
position and without indication of a fight.
u
The unicorn"-elephant fight is one of the most widely
spread motifs within the whole myth. Not only is it a standard
feature of many classical and Islamic authors, but it is also
to be found in the books of great Christian writers such as
Isidore of Seville
43
and Timothy of Gaza. 44 Through the
French-Norman poem of Guillaume le Clerc it was introduced
in the medieval bestiaries and their series of miniatures, 45 and
it is to be found in the lore of the Renaissance as, e.g., shown
by the poetry of Joshua Sylvester, by some tapestries, 47 and 46
by the inscription on Diirer's woodcut of the rhinoceros. 48

Through the Jesuit Ferdinand Verbiest it was in the second
half of the seventeenth century even introduced into Chinese
49
literature. It seems therefore appropriate to investigate
42
For the Tell Asmar seal see 104a, pp. 305-306 and fig. 108, a reference
kindly pointed out to me by D. S. Rice. For the Mohenjo-daro seal impression
see 179, vol. 2, p. 395; vol. 3, pi. 116, No. 14.
43
138, p. 435, paragraph 53.12.
44 2 paragraph
59> vol. 3, p. 297, 45.
45
271, pp. 33 and 35, pi. 156; 74, p. 17. The text of Guillaume le Clerc
(written in 1210) goes back to the Latin prose work Liber de bestiis ... of
Hugh de St. Victoire. Guillaume regards the unicorn as "si combatanz e si
hardie, qu' as olifanz prent atie," but in its fight with the elephant it uses its
claws and hoofs to rip the belly of its adversary (116, pp. 102 and 282-283).
46
242, pp. 217-218.
47 110, vol. (Flemish-German, ca. fig. 2 (Paris, mid-
2, fig. 148 1540) ; 13,
sixteenth century).
48
87, pp. 46 ff.
49
165, pp. 78-79.
No. 3 I. The Unicorn Ettinghausen 8.5
what the relationship between these two animals actually is.
In doing so this writer has to admit that he never witnessed

an encounter between the two animals. Therefore he is in a
not much better position to judge this problem than Cosmas

Indicopleustes, whose account is partly based on stuffed ani-
mals, or al-Jahiz, who merely evaluated literary reports with-
out having seen the animal. A modern writer has, of course,
a larger amount of first-hand evidence at his disposal, includ-
ing that of modern zoologists trained in the observation of
animals. However, it is astounding how even these well-
trained observers can come to different conclusions.
The best zoological opinion seems to be that the rhinoceros
50
is "as a rule a quiet, inoffensive animal." Its senses of hear-
ing and smell are extraordinarily acute, but its eyesight is so
poor that it is of little use. Owing to this condition the ani-
mal depends mainly on its auditory and olfactory senses to
warn it of potential danger. It is constantly on the alert and
51
a sudden sound or scent can send it into uncontrolled panic.
When the animal "gets one's wind ... it will immediately
go off at full speed in any direction ... if facing you it is
52
liable to run in your direction." This dash is actually more
of a flight than an attack and the animal seems to be unaware
of what it is heading for. To the untrained or frightened ob-
server this looks, of course, like a wanton and ferocious at-
50 Blanford adds to this statement the following very significant

48, p. 473.
passage: "the stories of its ferocity and its deadly enmity to the elephant
that were copied from the not very veracious pages of Captain Williamson's
Oriental Field Sports into European works on natural history, being fables."
(See also 240, vol. 5, p. 160.) Brehm (53, vol. 12, p. 608) states that "das
indische Nashorn ist . . . ausserordentlich bosartig," but he explains in another
passage that every rhinoceros is good-natured ("gutmiitig") when it first comes
across man and is not irritated, only to become vicious when aroused or when
it has had bad experiences. The same Brehm mentions at another place that
Indian rhinoceroses are "iiberaus gutmiitig und harmlos" (p. 614).
51
252, p. 3.
52 128, p. 11. Brehm mentions how a rhinoceros ran away from a single
dog, or as soon as it got the scent of a man (53, vol. 12, pp. 613-614). There
is also historical proof for the reluctance of the rhinoceros to accept a chal-
lenge. Babur mentions in his memoirs that during a hunt the animal would
not attack either man or horse and fled for 2 miles into the plains before it
could be killed (32, vol. 2, p. 451).

7
tack and just like the destructive disposition attributed to the

53
animal by popular belief. The impression of an attack is
further strengthened by the fact that the rhinoceros can trot

and gallop elephant which can only walk
in contrast to the
54
or "Only when wounded and driven about, the rhi-
shuffle.
noceros will sometimes charge home, though this is an ex-

55
ception." It will then tackle anything it can dimly see and
56
against which it rushes on in a straight line. His targets
can thus be trees, rocks, locomotives, trucks, the elephant
mount of a hunter,
57
and so on. When it attacks, it "is said
to use its sharp lower incisors (or as some think, lower
canines), much as a hog does." 58
On the other hand when
59
the animal is not frightened it can be easily tamed.
The elephant, being on the whole a rather timid, inoffen-
60
sive animal, tries to avoid the great Indian rhinoceros which,
at times, it seems to fear and from which it runs away. This
is said to apply equally to the wild elephants of Bengal and
61
the trained hunting animals, although there is, as we shall
presently see, other proof of the courage of the elephants.
There is no reason to assume that there is a deadly fight for
pasturage between them, as the classical writers believed. 62
53
128, p. 16.
54
48, pp. 474 and 465; 256, vol. i, pp. 40-41. Shortridge states that the
elephant "never attempts any pace beyond a shuffling kind of trot" (243, vol. 1,
p. 368).
55
48, p. 473-
56 and
53> vol. 12, pp. 611 614.
57 160.
148, p. 233 ; 240, p.
58 Blanford (48, pp. 473-474). To this zoologist was "shown in Cooch Behar,
a straight horizontal scar on one of the Maharaja's elephants just between
the feet . . . such a wound could not have been produced by the horn of the
rhinoceros." Cf. also 128, p. 3; 240, p. 160.
59
165, pp. 80-81, footnote 2; 252, p. 3.
60
48, p. 466.
61
38, pp. 14, 54, 91, 145; 240, p. 160. Shebbeare states about the great Indian
rhinoceros: "Wild elephants, though only too plentiful in the surrounding forest,
give the rhino and its haunts a wide berth" (241, p. 1229). The relationship
between the two animals in India seems to be the opposite from what it is in
Africa (see above, pp. 79-80).
62 This is also the opinion of Steier
(247, column 1786) and Wellmann (274,
column 2251).
There even exists a report of the peaceful encounter of a Sumatran rhinoceros
The attitude of the rhinoceros to smaller animals is quite

unpredictable. To quote a recent observer of the African
species: ". . . sometimes he feeds or drinks amidst a herd
of mixed species, apparently oblivious to their presence. On
other occasions he obviously wants the area all to himself,
and makes no bones about it." The same author tells that he
watched "a rhino charge into a herd of buffalo scattering them
in all directions, and on another time a female rhino and calf
sent a group of ten giant forest hogs scrambling away from
the waterhole. The rhinos were not content to drink until the
forest hogs had entirely abandoned the forest clearance in
which the muddy pool was located." 63 This attitude is exactly
the one which impressed itself so much on the Muslim mind
and which has found such frequent expression in the figural
arts.
The belief in the myth of the elephant-rhinoceros combat

was so strong that at certain times the animals were set against
each other. Thus when King Emmanuel of Portugal received,
in May 15 15, an Indian rhinoceros as a gift of Muzaffar,
King of Cambay, the first specimen to reach Europe since
Roman times, he matched it against an elephant. The fight
took place in Lisbon, on June third, 15 15. As described in a
letter by Valentin Ferdinand of Moravia, it proved to be a
disappointing affair, since as soon as the elephant noticed the
rhinoceros, it turned and with every sign of fear tried to
tail
get away; it finally broke some iron bars of a window and

64
managed to escape.
and an elephant which recalls the one given by Stott of his experience with an
African rhinoceros (see above, p. 80). Sincehave not been able to find such
I
a report on the Indian species, I am quoting, at least the passage about the
episode at a water hole in Burma:
and sure enough saw, not only
"I crept to the pool along an elephant path,
a double-horned Sumatran rhinoceros with a very
fine posterior horn, but also a
fine bull elephant with a very good pair of tusks. The latter was throwing
mud and water backwards over his body and between his forelegs. To cool
himself and to drive away the gadflies, the rhinoceros was standing alongside
the pool within ten yards of the elephant which seemed to take absolutely no
notice of its presence. A cock silver pheasant was also standing between the
two animals" (258, p. 149).
83
252, p. 4; 243, pp. 421-422.
64
104, PP- 33-41 ; 87, p. 50.
Even at the court of Lucknow, where the king arranged

all kinds of animal fights, the contest between a rhinoceros
and an elephant was not very exciting and altogether not easy
to effect since the two animals would attack each other only
when both were must. 65 The account given to us by Knighton
in his Private Life of an Eastern King discredits the unhappy
end in certain classical and all Muslim stories. At the same
time, it contradicts the medieval Muslim representation of
the fight according to which the karkadann approaches the
elephant from behind since it is assumed that the elephant is
fleeing from its enemy. Knighton's report states: "The ele-
phant approaches as usual with his trunk thrown up into the
air and head protruded; the rhinoceros either standing upon
his guard, or also advancing with lowered snout. The tusks
of the elephant sometimes pass on each side of the rhinoceros
harmlessly, while the huge head shoves the lighter animal
backwards. If the elephant's tusks trip up the rhinoceros, as
is sometimes the case, they are then plunged into him without
mercy; but more frequently the contest ends to the disadvan-

tage of the elephant, by the rhinoceros inserting his snout
between his antagonist's forelegs and partially ripping him
up ; the elephant belaboring all the time with his trunk, to a
certain extent uselessly, however. Prevented by his tusks, the
rhinoceros cannot get his snout far under the elephant's body,
so that the wound he inflicts is not generally a very severe
66
one."
05 This pathological condition results either in peculiar paroxysms of excite-
ment or in melancholy; it has often been described in the case of the elephant,
hut its causes are still little known. The usual explanation on a sexual basis is,
at least to a certain degree, questioned by many Western writers on the
subject. (See the comments on the most important literature in Edgerton's edi-
tion of the Mdtanga-lila, 197, pp. 29-38).
Al-Jahiz was the first Muslim writer to point out the importance of the
must condition, because he states that an elephant in heat is fiercer than a
karkadann (145, vol. 7, p. 24) Al-Mas'udi describes how the elephant attacks
;
its mahouts and guards and is then in such a wild state that even the karkadann
flees from it (184, vol. 3, pp. 57-58). This is, however, contradicted by
al-Nuwairi, who states that the karkadann attacks an elephant in heat, although
no other animal can withstand it at that time; the elephant draws back in fear
and his lust subsides (201, vol. 9, p. 316).
66
151, pp. 164-165. This book, first published in 1855, is based on information
The staging of such fights at the court of the kings of Oudh

suggests yet another possible origin of the myth of the ele-
phant-rhinoceros battle. In countries where people enjoy the
67
spectacle of animal combats, all possible combinations are
either actually set to test against each other or, at least,
thought of as possible rivals. In India both elephants and
rhinoceroses were available and there were affluent courts
staffed with the necessary men to arrange such fights. Al-
though deductions from the behavior of the rhinoceros dur-
ing hunting expeditions are probably of basic importance in
the formulation of the myth, actually staged (or just fancied)
combats may have given further substantiation to it.
In addition to these staged fights, the animals also encoun-
tered each other in the hunting field, when the elephant served
as a mount. Thus in 1525 Babur had the mahouts drive their
charges toward a roused rhinoceros which had broken cover
right in front of the elephants; the rhinoceros did not face
68
the challenge, but ran off in another direction. This must
have been a real disappointment for the Mughal emperor,
since he was interested in determining the outcome of an
encounter between the two monsters.
Another document demonstrating the elephant's bravery
is a miniature of the hunting Jahangir in the Sohn-Rethel col-
lection (pi. 33) ; it shows clearly that it would stand its ground
in face of a fierce attack and not run away.
What amounted to a fight took place during a viceregal
hunt in Assam in 1909. At that time the elephant was brought
to his knees while the rhinoceros rolled on the ground whence
it rose to its feet and disappeared into the jungle. 69
The only dissenting opinion is voiced by Bengt Berg who
reports that "it occurred several times on big hunting expedi-
tions in India that the belly of big hunting elephants was slit
open by the weapon of the aroused rhinoceros and the wounded
given to the author by an anonymous European courtier of King Nasir al-din

of Oudh (1827-1837). I owe this reference to Prof. K. A. C. Creswell.
67 "Beast fights of various sorts have been held in India from time imme-
morial" (47, p. 451).
08
32, vol. 2, p. 451.
69
53, vol. 12, p. 619.
go Freer Gallery of Art Occasional Papers Vol. i
elephant on the battle field." 70 Berg's statement is ap-

fell
parently not based on personal observation, and he may very

well have become a victim of the old myth.
There one more report, the earliest of them all,
is finally
which, however, does not indicate whether it was a staged fight
or an episode during a hunt. According to the version trans-

lated by Sachau, al-Blruni speaks of a young rhinoceros which
attacked an elephant, apparently frontally, because "the rhi-
noceros wounded with its horn the forefoot of the elephant
and threw it down on its face." Whatever the conditions may
have been under which the encounter took place, 71 the mode
of attack as given by al-Biruni is quite different from various
versions of the alleged fights as described by later Muslim
writers and then represented by the artists.
The conclusions to be drawn from these observations are
that certain elephants fear the rhinoceros and run away from
it, but this by no means always the case. The rhinoceros,
is
on the other hand, does not seem to be afraid of the ele-

phant, 72 but rather tries to avoid it. However, when disturbed,
especially by men, it will make a headlong dash which looks
like a charge; this may actually be a flight, though at times
it is a real attack. Only under special conditions, when the
rhinoceros and the elephant are must or infuriated are they
ready to fight. The outcome of these encounters is differ-
ently reported. seems that the frontal attack on the ele-

It
phant is not fatal. If the deaths claimed in the report by
Berg have actually occurred, an attack from the rear, i.e., on
a fleeing elephant, might be assumed. As to the origin of the
Muslim combat myth, it has not been possible to assemble
conclusive evidence, but the data so far gathered seem to
point to India.
70 91-92.
38, pp.
71 The text of a more extensive rendering of this encounter found in MarvazI
is unfortunately not well preserved in the two London manuscripts. Accord-
ing to this version the encounter was staged for Sultan Mahmud of Ghazna
(6, fol. 135a; 7, fol. 89a).
T-
According to Hubback "nothing except human scent seems to worry the
rhinoceros" (128, p. 7).
;
There are number of other features for which the Mus-

a
lim authors give the same kind of information as the writers
of the classical period. Usually, however, it is difficult to
prove whether there is an actual case of borrowing at hand.
Pliny and Solinus, 73 for instance, speak of the three-horned
oxen of India. This recalls al-Gharnati's words (repeated by
al-Damlri) which state that the rhinoceros has three horns,
one between its eyes and two above its ears. 74 Another unde-
cided issue is the position of the horn on the forehead of the
animal. Aristotle, 75 Pliny, 76 and Aelian 77 speak of this feature
however, when it occurs in Arab writings it may also be a
deduction from the position of the antlers and ordinary horns.
As to the equine features of the karkadann in Muslim repre-
sentations, it is unlikely that these can be accounted for by
an author like Aelian who speaks of a single-horned horse, 78
since in Muslim no such description is given. But
literature
there is at least one feature in Muslim texts which may pos-
sibly have been conditioned by classical authors. We are re-
ferring to the belief in different unicorns which were accepted
by even the outstanding writers of antiquity. 79 This conception
may have lingered on in Muslim times although the categories
of the Arab authors are different.
73 Pliny (209, vol. book

3, 8, pp. 54-55; 165, p. 104), Solinus, Collectanea, 52,
38 quoted by 74, p. 6.
74 The Chinese dictionary Er/i-ya with commentary by Kuo P'o (276-324)
having three horns, one on the head, one on
also defines the hsi rhinoceros as
the forehead, and the third on the nose (95 in 165, pp. 94 and 104). For a
three-horned animal, somewhat resembling a rhinoceros, although not quite
fulfilling the requirements of the Erh-ya, see the Six Dynasties figure in the
Eumorfopoulos collection (96, vol. 1, p. 17, No. 128, and pi. 17).
75 (speaking of the single-horned oryx and "Indian ass").
28, pp. 220-221
76 book
209, vol. 3, 8, pp. 56-57 (speaking of the unicorn).
77 104: book chapter 52 (speaking of the wild asses of India)
14, vol. 1, p. 4,
and 399: book 16, chapter 20 quoted by 74,

p. p. 11 (speaking of the animal
called Kaprd^wvos kartazonos by the Indians).

78
14, vol. 1, p. 76: book 3, chapter 41.
79 knew
of only two unicorns, but Aelian and Pliny
242, p. 34: "Aristotle
between them muster seven: the rhinoceros, the Indian ass, the oryx, the Indian
ox, the Indian horse, the bison, and the unicorn proper."
C. BIBLICAL AND EARLY CHRISTIAN WRITINGS

Biblical and early Christian influence on the lore of the
mythical unicorn is very limited. Al-Jahiz knew that the
Psalter mentions the unicorn and quotes these Biblical refer-
80
ences as a confirmation of the existence of the karkadann.
The most important contribution of the early Christian period
is the characterization of the harish and the story of its cap-
ture with the help of a young woman. The Muslim lore de-
rives obviously from that very influential and widespread col-
lection of allegorizing animal stories, called the Physiologus,
a work now thought have been composed about A.D. 370
to
in Caesarea, Palestine. 81
If one compares the Greek version
82
published by Sbordone with the earliest known Arabic ones,
by al-Tawhidi (see p. 60) one finds that the Physiologus text
is much shorter, although it is identical in the main features.
But one of the Syriac versions 83 is fairly close to that of al-

Tawhidi and it seems therefore that the influence came from
84
this direction. However, the Muslim accounts do not use
the story for allegorical purposes as the Christian versions
usually do, by likening the unicorn with Christ, by interpreting
80
145, vol. 7, p. 40.
81 2 75> PP 11 afi d 13-
82 208, pp. 78-80: "It is a small beast like a kid, but very fierce. A hunter
cannot approach it because of its great strength. It has one horn in the middle
of its head. How
then is it caught? They send out a pure and richly attired
virgin before and it jumps into her lap and the virgin gives the beast her
it
breast and takes him up to the palace to the king." See also 163, pp. 229 ff.
(quoted by 238, p. 181).
83 According Lauchert (163, p. 86) this is a "later" Syriac
to
207, p. 43.
version. Wellmann attributesto the seventh century (275, pp. 11-12), while
it
Ahrens stresses its importance for the understanding of the original Physiologus
(207, p. 9).
84 This Syriac version is closer to that of al-Tawhidi than the Christian-
Arabic version of the Physiologus published by Land (26, vol. 4, pp. 146-147).
This particular Arabic version lacks references to the size of the kid, its quiet-
ness and swiftness, its aggressiveness against animals, its "natural affection,"
the lack of milk in the breast of the virgin, the intoxication induced by the
sucking which is like that from wine, the tying with a rope and the motionless
behavior of the animal all these features are in al-Tawhidi and nearly all
;
of them in the Syriac text. Al-Tawhidi did not use a Syriac text directly, but
an Arabic translation entitled Naivddir al-hayaivdn (153, vol. 2, pp. 67 ff.,
footnote 1).
the single horn with the unity of Father and Son, and so on,
thus giving every actor and feature in the story a symbolical
meaning. 85 In Islam the passages about the harish are just
fanciful stories about a strange animal.
There is an early Christian feature of the unicorn story
which has its parallel in Islamic literature; however, it does
not represent a case of borrowing from a Christian source.
Cosmas Indicopleustes tells us that when the unicorn "finds
itself pursued by many hunters and on the point of being
caught, it springs up to the top of some precipice, whence it
throws itself down and in the descent turns a somersault so
that the horn sustains all the shock of the fall and it escapes
unhurt." 86 Although it has been noticed by a modern zoolo-
gist that the rhinoceros can jump down a sheer drop of 20
87
feet, the idea that it could break its fall with the help of its
horn is a mere tale. This feat has, however, been observed in
the case of the Persian wild goat (Capra aegragrus) which
can save itself after having made a false step by falling on
its horn. 88
This very incident is represented in the Mandfi'
manuscript of the Morgan Library in the chapter on the moun-
tain goat (buz-i kiihi) and illustrates the reference to this
89
feature in the text. There seems little doubt that in this
case it is Cosmas who borrowed from an account this trait
of the Persian wild goat or of an antelope which can perform
the same feat.
85 For this see 242, p. 48; 238, pp. 181 and 186; see also the allegorical inter-
pretation in the Christian-Arabic version of the story published by Land (26,
vol. 4, pp. 146-147).
86
76, p. 361.
87
128, p. 6.
88 McCrindle, the translator of Cosmas, tells us that the oryx
48, p. 503. is
said to have the same facility (76, p. 361, footnote 3), while Shepard mentions
also the ibex and the Rocky Mountain goat (242, p. 193).
89 182, vol. 2, pi. 25C. The "One of the wonderful traits
text, fol. 37b, states:
of the mountain goat is that it throws itself down from places that are about
one-hundred spears high and stands on its horns." The same iconographic fea-
ture occurs in Western Bestiary manuscripts as, for instance, The Pierpont
Morgan Library ms. 81, fol. 33 (ca. 1170). See also 40, fol. 9a (twelfth century)
and 41, fol. 190b (ca. 1310).
D. INDIAN INFLUENCES
It is evident that India, the homeland of the great Indian

rhinoceros, must have provided many features of the lore of
the Muslim unicorn. This is at once borne out by the fact
that the common names of the animal such as karkadann,
91
karg 90
al-bishan, and ganda, 92 and also sharav are all of
Sanskrit origin. Since at least one of these words from the
90
268, vol. 2, p. 820. Prof. B. Geiger (referring to Vullers' dictionary and
to Lagarde) confirms the old etymology which derives modern Persian karg
from Sanskrit khadga (1) sword, (2) horn of a rhinoceros, (3) rhinoceros;
and karkadan, Arabic karkadann, from khadga-dhenu, female rhinoceros (lit.
rhinoceros-cow). He does not accept the etymology presented by Ferrand
(98, vol. 2, p. 675) according to which karkadann derives from khadga-danta
"having sword- (like) teeth." His nonapproval of this etymology is based on
a number of reasons, especially on the fact that khadga means besides sword,
also rhinoceros (thus used in Vedic and classical Sanskrit texts) and according
to lexicographers also "the horn of the rhinoceros," and that, furthermore, the
characteristic feature of the animal is its horn and not its teeth.
Since the early days of Assyriology the Akkadian word kurkizannu has been
connected with karkadann and therefore has been translated as rhinoceros.
Dr. A. Leo Oppenheim informs me, however, that "the context (of kurkizannu)
clearly indicates the meaning 'young pig' and the correctness of this translation
has been borne out recently by such occurrence as e.g. 'one pig and its ten
kurkizannu.' The karkadann in Arabic, Syriac, Ethiopic (here: karkaned) etc.
can therefore not be connected with kurkizannu."

Dr. Cammann informs me that in Sung times the "Western" name of the
rhinoceros was known as ker-ka (second half of the eleventh century) ; this is
obviously a Chinese rendition of karg.
01 Thus in the Aklibdr al-Sln <v:a'l-Hind, but there are also misread versions
like al-nushan or al-nushan (184, vol. 1, p. 385, and vol. 3, p. 58), al-mushan
(Ibrahim b. Wasif-Shah), and jll*iJ | j\l*J| jLJ I in various manuscripts
of al-Nadim's Fihrist (192, vol. 2, p. 184). Reinaud already connected the
word with Sanskrit visana "horn." Al-Nadim (192, vol. 1, p. 349, and 98,
vol. 1, p. 130) and al-Jaihani as quoted by Marvazi (6, fol. 135b; 7, fol. 89b)
still call the horn of the rhinoceros (or a piece from it) by the name of
al-bishan although al-Nadim could have applied it also to the girdle decorated
with the horn pieces. Even when the term is given to the animal itself, such a
designation as ^JUlj jUu) I al-bishan al-mu'lam "the marked bishan" (Akhbar
al-Sin ica'l-Hind and al-Mas'udi) seems to refer to the designs in the horn.
Marvazi states also that the Chinese call it jl-ij bishan, which is borne out by
p'i-sha-na, the Chinese name for the horn; and in a Chinese Cham vocabulary
it is given as basan. (For a discussion of the nomenclature see 98, vol. 2, p. 675;
and 183, p. 82.)
92 JuS"" or Juf from Sanskrit ganda only given by al-Biruni and Marvazi.

Sanskritkarkadann or karg seems to have been current in

Iran during the late Sasanian period (see below, p. 150) we
can assume that Indian stories about the animal could have
been known in that country about that time.
Ithowever, difficult to establish actual parallels by quot-
is,
ing pertinent Indian passages since the rhinoceros, though oc-

curring as early as in the seals of the Indus valley civiliza-
93 94
tion and in Buddhist texts "hardly plays any role in the
cultural life, folklore or mythology of India. The allusions to
95
it in literary records are exceedingly sparse." Still quite a
number of Muslim borrowings from India can be established.
One such influence is only indirect, but it had great repercus-
sions on the Muslim lore; it came from the legend in the
Mahabharata about the young hermit Rsyasrnga. 96 He was
the son of the Rsi Vibhandaka and a gazelle and on his fore-
head thus grew a single horn. This caused him to be called
Rsyasrnga, "gazelle-horn," or according to certain Buddhist
versions of the legend, Ekasrnga, "unicorn." 97
When a
93 in, Nos. 341-347, Nos. 10-13; see copper

179, vol. 3, pi. pi. 116, a l so
tablets on pi. 117, No. 3. The clay figurines of the animal (pi. 97, Nos. 8-11)
made by children indicate the familiarity with and even the popularity of the
rhinoceros in the Mohenjo-daro civilization (179, vol. 1, p. 348; vol. 2, p. 387).
The animal Harappa (137, vol.
occurs also in 2, pi. 79, Nos. 74-79, pi. 91, Nos.
252-253) and Chanhu-daro (171, pi. 56, No. 8). For the inclusion of both
rhinoceros and elephant in the same figural representation, see above, pp. 83-84.
9i See the Khaggavisawasutta in
254, pp. 6-1 1. Prof. W. Norman Brown
does not translate the famous refrain as "but live as lives the rhinoceros
alone"; his rendition is "let him live alone, like the rhinoceros' horn" (60,
p. 218).
95
165, p. 112, footnote 2. Professor Geiger informs me that khadga, sword,
horn of rhinoceros, rhinoceros, and khadgin, rhinoceros, literally "having a
horn," do not occur often in the literature of India. In Sanskrit works are
some passages where the rhinoceros is mentioned in enumerations of wild ani-
mals or as having been killed by the arrows of a hero, while the law books
state that its meat is lawful for eating. Briggs (55, pp. 280-282) has gathered
references, nearly all of them modern, which illustrate the sacred position
of the rhinoceros in India. See also 118, p. 378.
96 The relationship to the Indian legend was first treated by Beal (11, p. 124,
footnote 2) and then further established by Luders (169, year 1897, pp. 114-115.
See also 165, pp. 110-112; and 238, pp. 181-182). The various versions of the
story are discussed in 169.
97
169, year 1897, p. 114.
drought struck the kingdom of Anga, the Brahmins advised

King Lomapada to bring the chaste Rsyasrnga into
the guilty
the country. The hermit is therefore visited by Santa, the
daughter of the king (or by courtesans, according to other
versions) whom the pious man imagines to be a penitent dis-
ciple. She arouses the hermit's love and then lures him to
her on a disguised float which is at once released to bring
cell
them Anga. Rain falls and the hermit marries the king's
to
daughter. This seems to be the ultimate prototype for the
Physiologns story of the capture of the unicorn with the help
of a young woman which, as we have seen, occurs again in the
accounts of the harish.
The legend of Rsyasrnga is also the raison d'etre of the
cervine type of unicorn. The genesis of its development is
quite clear. The first stage took place when the Indian legend
(or a derivative) was adapted for the purposes of the Physi-
ologus. The Christian version reappears then in the accounts
of the harish the earliest version of which so far traced is to
be found in al-Tawhidi (died after ioio). The harish ap-
parently soon became confused with the karkadann, since both
animals have a single horn. 98 As far as can be seen the wrong
identification appears for the first time in the Sihdh of al-
Jawhari (died between 1003 and 1010) which states that the
harish is the karkadann." Ibn Bukhtlshu', too, who wrote
only slightly later (he died after 1058) opens his account of
the karkadann with the remark that the Arabs call it harish
and then speaks of the ruse involving a maiden with which
the animal can be caught, a feature not quoted by al-Mustawfi
as being part of the account of the animal in the Sihdh. The
only remaining question is how the kidlike animal of the
just
Physiologns and of the Muslim texts came to resemble an
antelope or stag in the figural representations. In this connec-
y8 Laufer (165, p. 115) points out that the Sanskrit word vardhranasa means
both a rhinoceros and an old white goat-buck, confusion might
so that the
have been reinforced by semantic conditions, if they did not produce it entirely.
"The account is reproduced in al-Mustawfi (191, pp. 40-4 1 translation, )
p. 29). The eleventh-century dictionary of Asadi also states that the karkadann
has the shape of a goat (30, p. 105).
tion it is well to remember that the prototype of the harish,

the hermit of the Indian legend, was the son of a gazelle and
was originally called "gazelle-horned" (Rsyasrnga) and that
there was a general folkloristic belief in cervine unicorns in
the classical and postclassical periods,which possibly may go 1
back to Indian concepts. The myth of this type of animal

2
found figural expression also outside the Muslim world. This

fact was established by Laufer who referred to an illustration
<
of a cervine unicorn in the Chinese Cheng lei pen ts ao dated y
1208. 3 It therefore seems that the iconographic type goes

back to an earlier literary or figural model which was based
on the Rsyasrnga story or on general folkloristic notions about
a cervine animal. As stated before (p. 18) the analogy to
antelopes, the usual horn-bearing hunted animals, possibly
4
influenced the development of the type.
In other cases Muslim authors have specifically acknowl-
edged India as a source of the information they give in their
books. To this category belongs the story first told by al-Jahiz
that the karkadann prevents other animals from grazing in
its domain. The same applies to the yarn of the foetus stick-
ing its head out of the dam's womb prior to its birth, for
which both al-Jahiz and al-Biruni claim Indian origin. Al-
Qazwini gives an account of a caravan traveler who miracu-
1
Thus, e.g., Aristotle speaks of a unicorn gazelle (oryx). In the Cyranides,
a Greek work written between A.D. 227 and 400, it is said that the rhinoceros
is a quadruped resembling the stag, having a very large horn on its nose (165,
p. no).
2
This is assumed for the Cyranides by Laufer (165, p. 115).
3
165, p. 109. There is also a Chinese bronze figure of a cervine unicorn
in the A. Schoenlicht collection, which Visser attributes to the Sung period
or earlier (266, pp. 218-219) ; it is difficult to give a specific date to this piece
which might possibly be later than Sung.

4 The iconographic history of the sharav falls outside the scope of this paper
since it is although in Islam its lore became connected with
not a unicorn,
that of the Being the product of Indian imagination it can be
rhinoceros.
surmised that there are Indian pictures which could have influenced the
miniature of the sharav in the Berlin al-Qazwini manuscript (pi. 13, upper
left).That such animals exist in Indian art is shown by a seventeenth-century
Goanese bedspread in the Brooklyn Museum on which we find a bovine animal
with the characteristic four excrescences on the back (fig. 3; cf. 187, p. 130,
speaking of "strange spiny-backed creatures").
lously saved his companions from highway robbers by throw-

ing powdered rhinoceros horn mixed with dust on the brigands.
Since the man was an Indian, it can be assumed that the belief
in this quality of the horn was current in India. It may very
well be that some other alleged uses of the rhinoceros horn
come from the same country. This seems to be implied in the
statement of al-Qazwini and others about the effects of the
protuberance of the rhinoceros horn which is "only to be found
in the possession of the kings of India." At first sight this
statement looks, of course, suspect, since a branch of the rhi-
noceros horn can be nothing but a fanciful fabrication of the
mind. If one wants to attach any significance to al-Qazwini's
Fig. 3. Sharav (sarabha). Detail from an embroidered Indo-Portuguese bed-

spread. Goa, seventeenth century. Brooklyn Museum. (After 187, fig. 9.)
passage one would first have to explain how the notion of such
a branch could have arisen.
One possible solution of the riddle might be the psycho-
logical effect of seeing the famous horn for the first time. It
could very well be imagined that the Muslims must have been
amazed at its insignificant dimensions. Its small size was con-
trary to the traditional ideas about the horn so clearly demon-
strated by the pictures and it also fared very poorly when
compared with the tusks of the elephant. Since the horn was
in high repute it must have been difficult to imagine that this
was the whole of it and therefore the myth might have arisen
that this particular object in royal possession perhaps a cup
like the ones still kept in the Shosoin 5
was just a very power-
5 See also 246,
147, vol. z, pi. 31; vol. 7, pi. 8. p. 55.
:
ful branch of the main horn. This interpretation was all the
more natural because everybody was acquainted with branched
antlers of a stag with which it had been sometimes compared.
All this sounds reasonable enough to take the branch of the
horn out of the range of myths and into that of reality. Once
this has been established there seems no reason to doubt that
the effects of the "branch" in the possession of the kings of
India reflect Indian beliefs in the extraordinary qualities of
the horn itself. 6
It is a classical reference which points to India as the coun-
try of origin of a yarn first told in the Islamic period in the
Akhbdr al-Sin wal-Hind and then elaborated by al-Mas'udi,
namely that the karkadann has no articulation in its legs and
feet and therefore has to sleep erect, leaning against a tree.
Aelian suggests an Indian origin of this story by telling us
that the Indian unicorn called Kaprd&ovos kartazonos (which
corresponds to Sanskrit khadga-dhenu and Arabic karka-
dann) has no joints in its feet. 7 This motif seems to have
spread to the East, as far as China. In the T'ang period
rhinoceroses were said to have been captured with the help of
6 This high repute of the horn by classical writers such

in India first attested
as Ctesias, Philostratus and Aelian (14, vol. 1, p. 76:
(205, vol. 1, p. 234),
book 3, chapter 41, and p. 104: book 4, chapter 52) lasted throughout the
centuries. In the sixteenth century the Portuguese physician Amatus Lusitanus
writes
"The men of our Portuguese nation who have penetrated the interior of
India . . . say that its horn (i.e., of the unicorn) is greatly prized by the
Indian kings, and also those who have practiced medicine for some time in that
country and have then returned home say that in India there is no stronger
or more dependable antidote against poison than the horn of the unicorn" (242,
P- 143)-
Lately A. H. Godbey has suggested that the antidotal power of the rhinoceros
horn may not have been in Ctesias' original account, since neither Aristotle nor
Pliny mention this feature, though both were familiar with his text and used it
in their writings. It would thus appear to be a later interpolation of the text

which is preserved only through quotations in other authors. If this plausible
assumption should prove to be correct (and the lack of any reference to it in
the early Indian literature supports it), the belief in the magical virtue of the
horn during Roman times would probably go back to Chinese superstitions
which were imparted to the Western world when Roman traders exported the
horn for the Far Eastern market (109a, pp. 262-263, 277-278).
7
14, vol. 1, p. 399: book 16, chapter 20, quoted by 74, p. 11.
ioo Freer Gallery of Art Occasional Papers Vol. i
posts built of rotten timber which gave way as soon as the

animal leaned against them, leaving the jointless monsters
8
helpless on the ground. Such an alleged hunting custom is
also described in the West, but here it is the jointless elephant

which comes to grief by resting against a partly sawed tree. 9
In the Physiologus the story of the elephant without joints in
10
his legs is again connected with India, but Strabo speaks about
11
it in connection with the African species. It remains, of
course, a puzzle how such a myth could be connected with the
elephant whose habits were familiar to so many Indians. It
could perhaps be imagined as Laufer has thought
12
that
the story was first connected with the little-known rhinoceros,
13
since it is found in classical, Muslim, and Chinese sources.
Later on the myth possibly was transferred to the elephant
14
in regions and by people unfamiliar with the elephant, just
15
as Caesar attributed the jointless legs to the elk and Pliny
16
to the achlis.
There are, however, certain points which speak against
Laufer's theory. Thus, it seems to be a fact that the elephant
sleeps standing up,sometimes leaning against a support. He

assumes this position "not from any difficulty in lying at
length on the ground, but rather from the coincidence that
8
164, year 1913, pp. 361-362; 165, p. 146. In the Chinese accounts no direct
reference to India is made, only to "maritime people" living in a mountainous
country.
9
This myth goes back to Ctesias, just like that of the unicorn (274, column
2249). It was opposed by Aristotle who denied that the elephant sleeps stand-
ing and stated that it bends its legs and settles down (27, book 2, chapter 1:
p. 498a, lines 8-13. Cf. 275, p. 30).
10 The passage The
208, p. 130. is quoted in 164, year 1913, p. 362. motif
in the Physiologus has been traced back to Ctesias (275, p. 30).
11
253, vol. 7, pp. 324-325. Many classical, medieval, and renaissance ref-
erences to this elephant myth are given by Tennent (256, pp. 32-38). We
quote one from Troilus and Cress'ida: "... the elephant hath joints, but none
for courtesy: his legs are legs for necessity, not flexure" (act II, scene 3).
12
164, year 1913, p. 363.
13 See above, p. 99.
14 The jointless elephant occurs also in Muslim literature, e.g., in al-Mas'udi
(184, vol. 3, p. 8) and al-Mustawfi (191, pp. 32-33, translation, pp. 23-24).
15
66, pp. 352-353.
16 book Rackham thinks that the achlis perhaps
209, vol. 3, 8, pp. 30-31. is
the moose or reindeer and presumably a vernacular name.

No. 3 I. The Unicorn Ettinghausen ioi
the structure of his legs affords such support in a standing

position that reclining scarcely adds to his enjoyment of re-
pose. Elephants have been known for months to-
in captivity
gether to sleep without lying down." 17
On the other hand
the rhinoceros lies down on its side to sleep and it sits up with
its legs doubled under it or it rests on its haunches. This is
significant, although the animal has also been observed to rest

and sleep while standing. 18 All this points to the conclusion
that we do not necessarily have to accept Laufer's theory of
a priority of the rhinoceros myth which was later on trans-
ferred to the elephant and other animals. This is one of the
many cases where we still stumble in darkness and have yet to
bring definite proof to show from where and how the myth
was disseminated. Certain data speak for India but they are
not conclusive enough.
E. CHINESE INFLUENCES
It is unusual to find Chinese ideas incorporated in a Mus-

lim concept but the case of the rhinoceros was exceptionally
favorable for such a process. The Arabs were among the
chief exporters of rhinoceros horn to China. The Chinese
interest in horns with special designs must have been known
to the traders who were anxious to get the best possible prices.
They in turn told of the queer Chinese ideas when they were
back home and thus Chinese concepts found an echo in Mus-
lim writing. It is significant that it was a book revealing mer-
cantile relations with the East, the A khbdr al-Sin wa'l-Hind,
which apparently was the first to speak of the various patterns
on the horn. As has been shown by Laufer, writers of the
T'ang period already speak of the "likenesses to all things"
19
in high-grade rhinoceros horn, although the Chinese sources
speak more conservatively about grain or flower pattern, im-
20
pressions of beans, etc., when they mention specific motifs.
17
256, p. 39; 243, vol. i, p. 365.
18 See also 258, p. 146; 252,
263, p. 234. p. 3.
19 and p. 144: "The horn is filled with figures resem-
165, fig. 141, footnote r,
bling objects of nature." This statement dates from A.D. 860.

20
165, p. 141, footnote 1, p. 143.
8
io2 Freer Gallery of Art Occasional Papers Vol. i
However, Wang P'i-chih, a Chinese writer of the end of the

eleventh century, gives an elaborate repertory of these de-
signs which include not only plants and animals but also parts
of the dress, deities, scenery, buildings and so on, 21 so that it
is obvious that a rich lore had developed once the horn had
22 23
been used not only for official girdles, but also for cups,
knife handles, scepters, footrules, and other implements dur-
ing the T'ang period. 24 On the other hand, the lack of specific
25
descriptions of grain pattern in other natural products indi-
cates that the whole concept in Muslim literature is obviously
borrowed. 26 Since the idea is foreign to the unicorn lore of
21 (footnote 7 of preceding page). The Chinese author seems

165, p. 149 to
speak of natural pattern in the horn, but he indicates that these designs are
made more easily discernible by the work of craftsmen. He says: "When the
horn is completed into a carving, as if it were a veritable picture, it is highly
esteemed by the people."
22 The
earliest reference to the horn of the rhinoceros (hsi) in China so far
found from the year 311 B.C., that is from the end of the Chou dynasty. At
is
that time it was already regarded as an object of value and together with 300
boats, 500,000 arrows, and elephant tusks was sent as a gift by the King of
Yueh to the State of Wei. In a slightly later text it is indicated that objects
were made of rhinoceros horn and ivory (46, p. 328).
23 On the use of rhinoceros horn for the Chinese costume, see Laufer
(165,
pp. 142 ff., footnote 4). The horn used for girdles had to be of special quality,
owing to their official use, while that used for vessels could be of lower
standard. "If the specks are deep in color, the horn is suitable to be made into
plaques for girdle ornaments; if the specks are scattered here and there, and
light in color, the horn can be made only into bowls and dishes" (165, p. 141,
footnote and pp. 143-144. See also above, p. 54, footnote 8).
1,
24 Such objects are illustrated in 147, vol. 1, pis. 21, 24, and 31; vol. 3, pi. 8;
vol. _
6, pis. 18, 19, 20-23, 2 5 3 I ana 371 vol. 7, pis. 8, 51, 56, 58, and 59; vol. 10,
>
"
pi. 22; and vol. 11, pis. 46, 49-51, 54, and 55.
25 As far as am aware the descriptions
I are rather general and nonspecific
like that of Ahmad Tusi who at the end of the twelfth century states that the
bark of the khadhink tree (SjSs*) "shows pattern ... as if worked by a

Chinese painter," adding soon afterward that "it shows patternlike damascening
in steel" (134, p. 213). It seems characteristic that according to al-BIruni
the term khalanj does not only mean onyx, but also "all things which
show colored lines and figures such as cats, foxes, civet cats (?), giraffes, and
especially a certain wood of which, in the lands of the Turks, tables, cups,
drinking vessels, and similar objects are made" (134, p. 214).
26 It may, however, be possible that the alleged designs on the horn of the
rhinoceros induced Muslim artists to carve the outside of the much more com-
mon elephant tusks with various animals. This custom is first traceable in the
from all we know, apparently also to

the classical world and,
that of ancient India, the borrowing probably comes from
China, although Muslim writers, in some instances, may have
exaggerated Chinese ideas of the designs. 27
In a few other cases, it is not quite certain that ideas which
the Muslims thought to be of Chinese origin derived actually,
or at least exclusively, from that source. To this group be-
longs the lurid account of al-Gharnati (repeated by al-Damiri)
of how the kings of China tortured a person by having him
licked by the karkadann, a treatment which separated the
flesh from the bones. (The idea of a thorny tongue is as
we have seen also to be found in the story of the flight of
the sinad calf from its mother). It is true that a Sung pub-
lication speaks of the prickly surface of the tongue of the
rhinoceros, 28 but so does the Venetian Marco Polo who tells
of the spikes on the tongue of the Sumatran rhinoceros which

29
it uses when attacking a person. The feature is even to be
found in Pliny.
30
On the other hand the Emperor Humayun
31
is said to have told this story about wild yak. The story of
32
the prickly tongue may therefore have come to the Islamic
world from quarters closer than China.
In the report of al-Gharnati (repeated by al-Damiri) the
inlays of rhinoceros horn on thrones are not specifically called
a Chinese custom. Since they are mentioned together with the
alleged designs on the horn and the Chinese girdles, it is, how-
ever, not impossible that the Muslim writers had a Chinese
Fatimid period and therefore later than the first reports on the designs in the
rhinoceros horns, those in the Akhbdr al-Sin iva'l-Hind and al-Mas'udi.
27 footnote 7 (reference on p. 149).
165, pp. 148 ff.,
28 Su Sung, author of the T'u chlng fen ts'ao (165, fig. p. 140).
29 210, vol. 2, p. 285.
30 211, p. 340, footnote i. I have not been able to trace the passage in Pliny
to which the editors of the Marco Polo edition refer.
31 210, vol.
1, p. 277, footnote 3. For a reference to the prickly tongue of wild
oxen of India see 99, p. 283, footnote to p. 267.

32 Laufer explains the fable of the prickly tongue of the rhinoceros in the
following manner: "The animal mainly feeds on herbage and the alleged or
real observation of its inclination for brambles led to the conclusion that its
tongue must be thornproof and prickly" (165, pp. 140 ff., footnote 7, reference
on p. 141).
origin in mind. I have not been able to find a reference to such

a Chinese custom in the Sung literature, but there was a court
atelier during the Yuan period which amongst other things
turned out couches inlaid with rhinoceros horn and ivory. 33
China, however, influenced the Near East not only in the
conception of the karkadann and its horn, but also its iconog-
raphy. In spite of the distance between the centers of the two
civilizations, Chinese representations of unicorns had prob-
ably greater influence on the minds of Near Eastern artists
than those from any other country. In certain cases the Far
Eastern influence is obvious. There is first of all the karka-
dann on the title page of the al-Qazwini manuscript in the
Metropolitan Museum (pi. 17). It has the characteristic
shape of the ch'i-lin of the Ming dynasty with its split hoofs,
scales, and bushy tail. The only difference from the contem-
porary Ming examples would be the wings instead of the
flames, and, what is more important, the single horn instead
of two as we find them, for instance, on the Ming mandarin
31
squares. However, the literary references in the Chinese
encyclopedias of that period follow the older translations
35
and speak of a single horn, so that an illustration following
a literary source or a pre-Ming model might have influenced
the illuminator of the Persian title page.
A Chinese model is also the source for the miniature of the
shadhahvar in the al-Qazwini manuscript in the Princeton Uni-
versity Library (pi. 43, upper). Its lionlike body, tail, and
paws together with the use of a single horn seem to indicate
that the animal is derived from representations of the hsieh-
33 footnote 4 (reference on p. 143) The official in charge of this

165, p. 142, :
atelier was appointed in 1263, and there were 150 workmen active in it. In
view of the relationship between the rhinoceros horn and maritime ivory (which
will be discussed in the following section) it should be pointed out that narwhal
and walrus tusks were likewise used in the making of thrones. Old Russian
stories speak of precious chairs made of this material, there called "fish teeth"
(164, year 1913, quoting von Baer, 34, p. 10). The best known example
p. 338,
is, however, the Danish throne built of narwhal tusks, which was made in
1662-1665 (238, pp. 240-243, figs. 250 and 251).
3
*68, 4 b.
fig.
35 Kind information of Dr. Cammann.
36
chai. The main difference between the two animals is the
color, which in the Persian miniature is red instead of white.
If this change is not due to the fact that the Persian artist
arbitrarily changed the color, it might be surmised that the
ultimate prototype was a Chinese woodcut which showed the
animal only in black and white.
Far Eastern influence may also be assumed in the two
Shdh-ndmah miniatures of the Mongol period in the Myers
collection (pi. 24). The basis for this supposition is the fact
that the Chinese knew of a lionlike creature with a single horn
(fig. 4), and as we have seen in our discussion of Indian infiu-
Fig. 4. One-horned leonine monster from a Chinese bronze mirror. T'ang period
(618-907). Freer Gallery of Art. (Drawn by Mrs. Eleanor M. Jordan.)
ences, China also knew from the early thirteenth century on,
at least, the iconographic type of a single-horned cervine rhi-
noceros. Thus there is a certain possibility that the prototypes
of these monsters came from the Far East. This is made
more likely by the fact that these animals show flames above
their front legs, a typical Chinese feature which occurs on
dragons and on some other animals in the pre-Ming period. 37
But while there is no doubt about the Far Eastern origin of
36
68, fig. 4 a.
37 The Freer Gallery of Art has a T'ang mirror with
a large dragon with
flames emanating from above the front legs and another from the
(No. 38.8)
same period showing two horses, one apparently even horned, with flames above
the front legs (No. 44.5; pi. 48, upper). Other mirrors from the Sui and T'ang
periods with such flame-endowed horses in 261, part 2, vol. 2, pis. 117 and
129. Dragons of the Sung period commonly show flames.
this flame motif, both the shape of a lion and of an antelope

for a karg could be explained by certain details given in the
Shdh-ndmah and other Muslim texts.
The ch'i-lin, the hsieh-chai, and the flame motif should be
regarded as "chinoiserie" in Muslim art. Fairly common as
these and some other Chinese mythical creatures are in cer-
tain periods, they still represent an unassimilated foreign ele-
ment. There is, however, one Chinese animal which became
thoroughly integrated into Muslim art, namely that of the
buffalo or oxlike type of karkadann. Such a statement might
seem surprising as the many references in Muslim texts would
seem to point toward an indigenous concept of this sort.
However, a confrontation of the oldest example, that in the
Manafi* manuscript of the Morgan Library, with a Chinese
painting will quickly dispel any doubts about the Far Eastern
origin of the type.
For comparison we have selected a Buddhist painting in
the Freer Gallery, showing "Lohans moving through forests
and sea" (pi. 12). It dates from the Ming period, 38 and is
thus centuries later than the Mongol Mandfi' miniature, but
its rhinoceros corresponds to the descriptions in T'ang and
Sung and may thus be regarded as representative
literature
for the earlier periods. The first figure on this scroll is a lohan
riding on a buffalolike animal carrying on its forehead a
39
single large horn bent forward. There is no doubt that this
bovine animal is the one-horned variety of the hsi rhinoceros
40
of the Chinese which is described in the fourth-century edi-
3S Freer Gallery of Art, No. 19.174. Siren attributes the scroll to the late
Sung period and Fan-lung (a priest) (244, pp. 57-58, pi. 93). A. G. Wenley
to
attributes the painting, however, to the late Ming period; he thinks that the
painting could only be a very free copy of Fan-lung's work.
39 That a rhinoceros with a forward-turned horn goes back to a much earlier
period, is demonstrated by the figures on a kettle attributed to the Shang

period (165, fig. 18).
40 There is also proof that the hsi-rhinoceros and the ox came to be regarded
as identical animals so that in legends the names of the two were exchanged
(90, pp. 328-329). In this connection it is also significant that although from
post-Han times down to the present both ssu and hsi mean rhinoceros, the first-
named animal was originally a wild bovine of large size and formidable
nature, probably the gaur, while hsi stood always for rhinoceros (46, pp. 322-
33o).
tion of the Erh-ya Dictionary in the following terms: ". . .
the hsi resembles a water buffalo with a pig's head, has a large
stomach and short have three hoofs. Its color
legs. Its legs
is black. It has three horns, but there

is also a one-horned
type."
41
When one now compares this Chinese animal with
the Persian karkadann, one finds the same large penetrating
eyes, long ears, wide nose, the hump, and the feet.
42
How
slavishly the Persian artist followed the Chinese prototype is
particularly shown by the painting of the horn with its pecu-

liar curve. The artist first placed it farther down on the head,
with the curve going backward, thus coming close to the actual
position of the horn of which he seems to have had some idea.
But another look at his painted model of Chinese inspiration
must have shown him his "mistake," which he hurried to make
good, although traces of his first endeavor are still noticeable
under the leaves of the surrounding shrubbery. The Chinese
painting even explains a peculiar feature in the Near Eastern
miniature, namely the zebralike stripes between the front legs
and the head. They are the misunderstood skin folds of the
neck which, like the folds in the hides of the famous "ele-
phants" in the same manuscript, 43 were turned into a pattern
by the artist; and since the painting from which he worked
apparently showed the ribs of the animals, he converted these,
too, into stripes. All these features reveal an integrated re-
44
creation after a foreign model.
It may very well be that the Chinese model not only influ-
enced the painters, but also the writer of the Persian version
of the Mandfi'. He states that the karkadann closely resem-
bles a buffalo, that its neck is strong and long, that the single
41 The translation was kindly supplied by Dr. Cammann.

Laufer (165, p. 89) had already pointed to the similarity of the Morgan
42
miniature to a woodcut illustrating the Erh-ya. Since he compared it also

with the representation of a bovine animal on an Assyrian obelisk (which had
erroneously been associated with a rhinoceros), he diminished the importance
of his discovery.
43 820A.
154, vol. 5, pi.
44 Until recently it was generally believed that group of miniatures in
the
the first section of the Mandfi' -i hayavdn to which karkadann belongs was
the
painted in pure Abbasid style and thus devoid of any Far Eastern influence.
io8 Freer Gallery of Art Occasional Papers Vol. i
horn on the forehead is crooked and near the eye and that
the tail and hoofs are similar to those of a buffalo. All these
details are very specific and since they tally better with pic-
tures of the bovine type of Far Eastern extraction than with
the real animal, it could be assumed that such pictorial records
are reflected in the text.
In several representations the Near Eastern artists did not
give the horn on the forehead of the bovine rhinoceros the
forward curve which we found in the Mandfi' miniature and
in the Ming scroll of the Freer Gallery. They replaced it by

a long straight horn. This was due to another outside influ-
ence of which we
have to speak in the next chapter.
shall
What has been said regarding the origin of the iconography
of the single-horned bovine karkadann in Muslim painting
may also apply to the double-horned variety which shows, as
we have seen, a horn on the nose and a longer one on the
forehead of the animal (pi. 15, right). The same type of
rhinoceros is found on a Chinese mirror said to be of the
45
T'ang period (fig. 5 ) . The type persisted throughout the
centuries, as demonstrated by a Ming mandarin square, for-
merly Mrs. Krenz, which shows the same
in the collection of
spotted bovine rhinoceros with two horns which we find in
46
the Princeton miniatures of the karkadann and the sinad.
In spite of this close parallel between the double-horned rhi-
noceroses in Far and Near Eastern representations, a case
for a different explanation of the Muslim version can also
be put forth. The double-horned variety seems to appear in
the Near East only after the correct position and the proper
shape of the horn in the single-horned variety has become
known and has been applied on the nose of the animal in a
number of miniatures. Even at this progressive stage the
artist followed, however, traditional patterns of thought and
design as we noticed, for instance, in the case of the antelope-
karg in the Shdh-namah of 16 05 -1608 (1014-1016) in the
45 Figure after 72, chapter 40, p. 56A; see also 261, part vol. 2, pi. 148.
5 is 2,
Kind information of Dr. Cammann.

46 Dr. Cammann kindly informed me about this textile, the present where-
abouts of which is unknown.
Metropolitan Museum. Could it not be assumed, therefore,

that the double-horned variety is nothing but the traditional
bovine animal with the straight horn on its forehead now
supplemented by an additional nasal horn? In this case we
would not have to assume a Chinese prototype for the two-
horned version of the karkadann.
Fig. 5. Chinese mirror. T'ang period (618-907). (After 72, chap. 40, p. 56.)
F. THE LATE MEDIEVAL WORLD

After having surveyed the pertinent classical and Oriental
sources, there remains now the sole task of exploring a possi-
ble between Muslim and medieval European
interrelation
writings on the unicorn. Such an investigation is important
because in the West it is generally assumed that the extraordi-
nary reputation of the horn of the unicorn as a detector of
poison was due to the influence of Arab doctors. 47 Their writ-
242, pp. 140-142; 238, pp. 189 and 200.

I IO Freer Gallery of Art Occasional Papers Vol. i
ings are supposed to have caused the rise of the new belief,
of which there is no trace in the medieval world before the
48
thirteenth century.
When one wants to investigate this Arab influence one
naturally starts with the karkadann which is the unicorn par
excellence in the Muslim world. In investigating a possible
relationship one is, at first sight, permitted to disregard the
great difference in physical appearance between the short,
curved rhinoceros horn and the long, straight tusk of the
narwhal which, to the late medieval mind in the West, was
49
the true horn of the unicorn. After all, the representations
from various Islamic countries seem to indicate that the Mus-
lim usually regarded the horn of the rhinoceros as long and
straight. There is, however, the surprising fact that with the
exception of al-Qazwini in the thirteenth century, his some-
what later follower, al-Mustawfi, and the latter's contem-
porary, Ibn al-Wardi, no other Muslim author consulted here
speaks of poison in connection with the horn of the karkadann.
In the eleventh century, Ibn Bukhtishu' deals with the medici-
nal uses of the animal but he does not mention the horn or,
for that matter, any other part of the body, as an antidote
against poison; nor does the late thirteenth-century Persian
zoological pharmacopoeia, the Mancifi', speak of it. Al-
48
Hildegard of Bingen (1098-1179), though speaking of the medicinal quali-
ties of various parts of the unicorn, does not mention the alleged virtues of
the horn. Albertus Magnus (1 193-1280) ''makes little of the horn's magical
virtues and thinks they should be investigated further." Pietro d'Abano (ca.
1250-1318) however, "during his exploration of Arabic lore acquired a firm faith
in the alicorn and indeed if one were asked to name a single writer to
. . .
whom the European belief might be attributed with least exaggeration, one
could not do better than to choose this Peter" (242, pp. 120-121).
49 Even European mind the rhinoceros horn often took the place of
in the
the true "alicorn" ornarwhal tusk. (Alicorn from Italian alicorno and French
licorne is used by Shepard to avoid cacophony [242, p. 101, footnote 1, and
pp. 141-142] and it seems a good term to use.) The tip of a rhinoceros horn,
presented by the prior and the brothers of the Monastery of St. Mary of
Guadelupe, Spain, to Pope Gregory XIV, in 1590, is said to have been used
in powdered form, as medication in the pontiff's last illness. The horn with
its fine case (of which I learned through the kindness of G. Schoenberger)
is now in the American Museum of Natural History, New York (168, p. 532).
Other such horns are mentioned by Shepard (242, p. 133).

1
No. 3 I. The Unicorn Ettinghausen i 1
Nuwairi mentions the poisonous quality of the horn only as

an explanation for the death of the gored elephant without
saying anything about the possibility that it could be used as
an antidote for other poisonous substances, especially in hu-
man food. Even al-Damiri, the most extensive writer on the
subject, givesno information on this point. Furthermore, al-
Qazwini does not seem to stress the antidotal property; he
merely mentions it
after having told his readers that knife
handles are made from the horn and without alluding to a
possible connection between the two uses. Finally, neither al-
Qazwini nor al-Mustawfi mention the "sweating" of the
horn, which in the case of the alicorn is supposed to have indi-
cated the presence of poison. 50 Thus the earliest Muslim
source for the antidotal quality of the karkadann horn is
almost contemporary with similar references to the alicorn in
European literature, and it lacks a feature of possible impor-

tance, the sweating. This makes it imperative to find out first
how al-Qazwim came to speak of the horn's antidotal effect

and then whether it is actually the karkadann which could
have given rise to the late medieval myth of the alicorn's
medicinal quality.
One might be tempted to assume, on first thought, that the
sudden rise of the belief in the antidotal qualities of the karka-
dann horn was due to Indian or Chinese influence, which in
other instances proved so important. When one examines the
al-Qazwim text closely one notices, however, that its power
to nullify poison applies to the horn as a whole and not to
the more effective "protuberance in the possession of the kings
of India." Hence, in the Arabic text this specific virtue is not
connected with beliefs which can be traced back to India. Since
relevant passages in Sanskrit literature seem also to be lack-
51
ing, the sources for an Indian origin of the horn's antidotal
50 The history of the "sweating" of the alicorn as far as

242, p. 119. is, I
know, not yet explored, so that I am unable to state how early this feature occurs
in the West.
51 Pelliot suggested that the antitoxic quality of the rhinoceros horn in India
may be due to a popular etymology which connected visana "horn" with visa
"poison" (98, vol. 2, p. 675). If this should prove to be correct, it may be
nature are the classical accounts of Ctesias, Philostratus, and

Aelian which the miraculous virtue is associated with cups
in
made Drinking vessels made of the horn are

of the horn.
hardly ever mentioned in Muslim texts, 52 nor have medieval
cups to my knowledge turned up in Muslim collections.
The same objection applies to the assumption of a Chinese
influence, since the Chinese likewise stressed the use of cups.
Indigenous sources, however, leave no doubt that from early
times on the Chinese attributed an antidotal virtue to the
horn. 53 Arabic and Persian texts give as little indication of a
regarded as the reason why the belief seems to have existed for a very long
time without having found expression in the higher form of Sanskrit literature.
52 I have come across only two cases from past centuries: The royal vessels
made of rhinoceros horn of which Ibn Fadlan speaks are called
taifuriyat (134, p. 34), a term which Zeki Validi translates as "grosse Prasen-
(134, p. 77). In view of the shape of the horn, this can hardly
tierteller"
be the proper interpretation of the Arabic word. Dozy renders it, however,
as "plat creux et profond" (88, vol. 2, p. 48) which together with other dic-
tionary definitions seems to establish "deep vessel" as its basic meaning. This
term could thus very well have been applied to a drinking cup made of rhi-
noceros horn.
The second reference is to be found in Babur's memoirs. He mentions a boat-
shaped drinking vessel made of the horn, but he does not speak of its anti-
dotal property (32, vol. 2, p. 389). In modern times Reinhart speaks of a bowl
of rhinoceros horn made in India which had the alleged antidotal quality
(219). Brehm (53, vol. 12, p. 623) also speaks of such drinking vessels.
Meyerhof mentions a cup, a bowl, and a bottle offered to him in Cairo; of
these only the last-named object seems to have been regarded as active in
detecting poison (98, vol. 2, p. 680).
53
246, p. 55. The horn was also used for stirring food, since, owing to this
action, poisoned food started to foam: "The horn is a safe guide to tell the
presence of poison: when poisonous medicines of liquid form are stirred with
a horn, a white foam will bubble up, and no other test is necessary; when
non-poisonous substances are stirred with it, no foam will rise." (165, p. 138,
quoting Li Shih-chen, a fourth-century authority.)
As to the origin of the belief Laufer has this to say: "The Taoist adept and
writer Ko Hung (died in A.D. 330 at the age of eighty-one) is apparently
the first Chinese to speak of the poison detecting quality of the horn, which on
contact causes liquid poison to foam. This property is accounted for by the fact
that the animal while alive feeds particularly on poisonous plants and trees
provided with thorns and brambles." This poisonous food is then thought
to affect the horn and according to the principle that poison cures poison the
horn becomes an efficient antidote (164, 1913, p. 323; 165, pp. 75, 138-139,
and 154, footnote).
No. 3 I. The Unicorn Ettinghausen H3
Chinese source for this particular belief as they do of an

Indian one.
When all these facts are considered one can only say that
Indian or Chinese conceptions could have affected the Muslim
mind. On the other hand, it seems unlikely that such influences
and such influences alone suddenly caused al-Qazwini to in-
clude the information about the antidotal quality of the horn
in his chapter on the karkadann. One is, therefore, forced to
a more plausible solution of the riddle.
One might be inclined to think that classical literature could
qualify as a possible source for the myth, because from the
time of Ctesias the horn of the rhinoceros (or whatever the
animal might have been called) had the reputation of pro-
54
tecting against poison. Yet if classical writing were to be
held responsible for this myth, traces of it would have been
found at earlier times (for instance in the text of Marvazi)
when the influence of classical authors was stronger than in the
time of al-Qazwini. It seems, therefore, that the solution of
the whole question does not lie in this direction. We have to
look for it elsewhere and with other methods.
In the course of this investigation, it has been found at times
that in an ancient source two different animals were thought
to be identical or were at least confused in the minds of the
55
people. This was usually the case when these animals be-
longed to little-known species and had one or more character-
istic traits in common. This not infrequent occurrence might
give us the required clue. We therefore have to investigate

whether or not there is a rare animal sharing certain features
with the rhinoceros which possesses a horn or horns endowed
with antidotal virtues. From such an inquiry we find that a
substance often thought to be a horn does exist, although the
animal from which it derives is mysterious and not properly
identified.
54 The classical writers besides Ctesias are Philostratus (Apollonius of Tyana)

and Aelian and 242, pp. 27, 35, and 39). As pointed out
(see 74, pp. 8-11,
above, the passage in Ctesias may, however, be a later interpolation based
on Chinese beliefs which trade brought to the West (see p. 99, footnote 6).
55 Karkadann and sharav; karkadann and harish; karkadann and sinad.
There are many such cases.

H4 Freer Gallery of Art Occasional Papers Vol. i
Al-Biruni tells or khutuw (jp^-) 56

us that the khutu (^)
is highly treasured, especially by the Chinese and the Turks
of the East, because it is believed that this animal substance
like the bezoar sweats when it approaches poison. 57 About
three centuries later in 1339 al-Mustawfi speaks of the
same miraculous properties but no longer restricts the belief
to the Chinese and states that poison has no effect on anyone
who carries it since he can see it sweating and be warned. 53
Two other authors, al-Akfani (died in 1348/749) 59 and al-
Ghaffari (wrote in 151 1/9 17), 60 likewise testified to the same
quality, al-Akfani furthermore adding that the khutu was a
favorite with the Turks and Chinese. This particular horn,
however, did not always have a reputation for possessing anti-
dotal properties, because the Hudud al-Alam written in 982
(372) and thus our earliest source mentions only that knife
handles were made of this material. 61 This earlier use as
knife handles which may have continued when the antidotal
50
The spelling khutu is found in the Hudud al-'Alam (129, pp. 62, 84., 92, 94,
96, and 97), MarvazI (1S3, p. 5, line 10, and p. 11, line 7) and al-Mustawfi
(191, Persian text, p. 20) and thus seems to be the Persian version. Khutuw
is the form chosen by M. J. de Goeje in his editions of al-Istakhri (139, p. 289,
line 1) and Ibn Hawqal (135, p. 337, line 14; see also III, p. 222), and it is
also used by E. Wiedemann for his translation of the pertinent passages from
al-BIruni (277, p. 353) it apparently represents the Arabic form.
;
r>7
Al-Biriini's Kitdb al-jamdhir fi ma'rifat al-jaivdJiir was not available to
E. Wiedemann when he translated the passage from this text (277, pp. 346-
347). He had to use al-Khazini's Kitdb mizdn al-hikma, written in 1121 (515),
which preserves a section of al-Biriini's work (277, pp. 353-355, also quoted in
164, 1913, pp. 315-316).
After this manuscript had been finished, F. Rosenthal kindly referred me to
the original passage which is new available in the Hyderabad edition of
1355 H. (45a, pp. 208-210). The two texts are not identical; they vary a
great deal, even in subject matter. An account of the special features of

al-Biriini's original text is given on pp. 138 fr.
58 translation
191, p. 20, p. 14.
59 footnote also quoted in 164, 1913, pp. 316-317, and comments
277, p. 354, 2,
in 164, 1916, pp. 380 ff. The work is apparently the one listed as No. 5 in 57,
vol. 2, p. 137, and suppl. 2, p. 169.

60
277, pp. 357-358 quoting G. Jacob; the passage is reproduced by 164,
I 9 I 3> P 3*7- The original texts of al-Akfani and al-Ghaffarl were not avail-
able to me.
61
129, pp. 84 and 96. On p. 94 the khutu is called a horn; al-Mustawfi refers
to it as the animal itself.
.
No. 3 I. The Unicorn Ettinghausen ii5
virtues became more widespread, links it with the horn of

the karkadann which, according to al-Birum, Marvazi, al-
Qazwini, the Mandfi' and al-Mustawfi was also used for this
'
,
purpose. 62 Both horns were also greatly in demand and there-

fore sold at high prices. Even as late as the sixteenth century
the khutu horn had a great reputation since al-Ghaffari tells
us that the padishah, apparently the Sultan of Turkey, bought
it at a high price and that its fame reached as far as China
and the Maghreb.
Additional remarks in the Muslim sources throw further
light on the physical character of the horn without giving any
obvious clues to its identification. The color is said to range
from white, yellow, green, and red to gray and black. 63 The
curved specimens are the highest in price (al-Biruni). The
price for a young horn is higher than for an old one (al-
Mustawfi) because the young ones are fresh and firm while
later on they become dark-colored and soft (al-Ghaffari)
There is no doubt that the animal from which the khutii
horns came is even more mysterious than the karkadann. I
have not found a single writer who states that he had ever
seen the animal or quoted someone else who was more fortu-
nate than he in this respect. It is therefore not astonishing
that even the zoological class to which the animal bearing
these horns belonged is uncertain. Al-Biruni quoted anony-
mous was the bone from the
authorities to the effect that it
head of a bull; Tusi thought it to be a snake bone, and al-

Akfani attributed it to a bird. The most important of these
identifications is that given by al-Biruni, as it is the oldest
62 The use of rhinoceros horn for the hilts of Egyptian swords is reported as
late as the early nineteenth century Burckhardt (98, vol. 2, p. 680).
by J. L.
Since the material has very fine coloring and can be beautifully polished, it is
well fitted for the purpose quite apart from other intrinsic values. About the
use of khutu horn for knife handles see below, pp. 117, 121 ff., 126-127.
63 Al-Biruni states: "Its best quality is the one passing from yellow into
green, next comes one like camphor, then the white one, the one colored like
the sun, then one passing into gray." Al-Mustawfi says that it is yellow in
color. Al-Akfani quotes the Ikhwan al-Raziyan, the two Razi brothers, that
it "changes from yellow into red. Then comes the apricot-colored one, then
that passing into a dust color and down to black." According to al-Ghaffari its
"color is yellow and the yellow inclines to red."
n6 Freer Gallery of Art Occasional Papers Vol. i
and therefore probably closest to the early forms of the myth.

Here we notice again how the lore of the khutu and that of
the karkadann could become intermingled, since in either case
the horn is said to belong to a bovine animal. The two other
identifications given above are due to later importations of
Asiatic myths from farther east; 64 they do not seem to be of
any consequences for the growth of the main conceptions of
the khutu horn and reveal only the general mystification about
its true origin.
The information about
the countries from which the horn is
supposed to come is more specific than that on the animal itself,
but this does not help to make up for the general uncer-
tainty. The Hudi'id al-Alam lists the following regions as
exporting the horns: The Tulas mountains southwest of Lake
65 66 67
Balkhash, China, Tibet, the Turkish Toghuzghuz country
68
in eastern T'ien-shan, and finally the Kirghiz country which
supplied large quantities. 69 According to al-Istakhri the horn
64 Laufer also has shown that the
(164, 1916, pp. 382-384; 166, p. 566)
khutu horn, said to come from the forehead of a large bird, is actually the
enormous bill of the buceros or hornbill which is carved by the Chinese and
Borneans. According to Malayan beliefs shavings from it become blood-red
when placed in a poison dish. There is also the bird al-khathaq (JpiM) of
"China, Babylon, and the lands of the Turks." Al-Damiri says of it that it dies
in the presence of poison and that people make knife handles from its bones
since it sweats in the presence of poison (77, vol. 1, p. 251; 78, vol. 1, p. 667,
quoted in 164, 1916, p. 388, footnote 1). References to the swooning or dying
of birds are a regular feature in Indian literature (70). Disregarding diacriti-
cal points, khathaq is written in the same way as chatuq, of which we have
to say more later on (see pp. 122 ff.).
There are altogether seven different zoological definitions for khutu in Vullers,
Lexicon (268, vol. 1, p. 659). One of them states that it is a thousand-year-
old snake, an explanation which goes back to the Liao Annals (166, p. 566).
05
129, p. 62.
CG
129, p. 84.
07
129, p. 92.
68
129, p. 94.
69 and The Kirghiz country also quoted by al-BIruni
129, pp. 96 97. is
though with certain doubts. The Hudud (129, p. 97) mentions, however, that
a certain clan of the Kirghiz hunts not only for furs and musk, but also for
the khutu horns. The importance of the Kirghiz region as a major source
of the khutu is corroborated by a Chinese source. According to the T'ang
Annals the ku-tu is mentioned as one of the wild animals of the country of
the Kie-kia-se; i.e., Kirghiz (Laufer, 164, 1916, pp. 370-371).
;
is exported from Saghaniyan (the region north of Tirmidh

on the Oxus), 70 while al-Ghaffari mentions the lands of the
Berbers and Turkestan as the habitat of the animal.
It is not surprising when, in this nebulous land of uncer-
tainty, yet another horn appears to add further confusion.
Al-Biruni speaks, in connection with the regular khutu, of the
tooth of a fish 1 cubit long which the Volga Bulgars bring
from the northern seas. 71 In Mecca, where it is imported, it is
regarded as "white khutu." The tooth is so much in demand
that the Egyptians "who crave it" paid two hundred times its
value. Al-BirunT does not give the reason for this extraordi-
nary popularity. When he states that white knife handles are
sawed out of the tooth he obviously does not tell the whole
story. Since it is called the "white khutu" it must share
another more specific and rare quality with the colored khutu,
not just the prosaic function of serving as a material for knife
handles. The most obvious connection, and the one which
would easily explain the high price for both varieties, is that
the tooth has the same poison-detecting faculty as the ordinary
khutu. It will be our task to substantiate this assumption.
Unfortunately the description of the physical character of
the "white khutu" is very meager. There is, for instance, no
indication whether the tooth is straight or curved. In other
respects, al Biruni's report is more specific but even so not
very clear. According to him one is able to distinguish the
"white khutu" from ivory, since he points out that the vari-
ous designs displayed on it give it the appearance of
72
wriggling.
There remains now only the question of what the khutu
horn actually is; and it is here that we run into great if not
70 Al-Istakhri (139, vol. Ibn Hawqal
1, p. 289, line 1) ; (135, p. 337, line 14).
If Pelliot's emendation of a passage of Abu Dulaf Mis'ar b. Muhalhil (of about
940 [329]) is correct, still among the Chigil
another place of origin would be
in this text it is cow (98, vol. 1, p. 211; vol. 2, p. 679). About
described as a
the Chigil see Minorsky's comments in 129, pp. 297-300 and maps V and VI.
71 2 77> P- 354; 260, p. 50.
72 The German translation of Wiedemann uses the word "durcheinan-
derkriechen." Al-Ghaffari states that "designs are displayed in the khutu
as in damascening."
1 1 8 Freer Gallery of Art Occasional Papers Vol. i
insurmountable difficulties. This is not the first time that a

scientific searchlight has been focused on these mysterious
horns. Just before and after World War I they were dis-
cussed by a number of Orientalists, who proposed rhinoceros
horn, mammoth and narwhal teeth as
ivory, walrus tusks,
possible solutions. Rhinoceros horn, as suggested by several
German scholars, 73 seems to be out of the question as the origi-
nal meaning because al-Biruni would probably have recognized
it as such and he would have spoken of its antidotal, qualities
in his account of the rhinoceros.

74
The Hudud al-Alam, too,
distinguishes between the countries which supply rhinoceros
horns and those which export khutu horns. Furthermore, it
should be recalled that al-Biruni states that the curved khutu
demands the highest price, which presupposes that there is
also a straight variety. This obviously does not apply to the
horn of the rhinoceros. Finally, it seems most unlikely that
the Egyptians, who were familiar with the horn of the African
rhinoceros, would have paid exorbitant prices for it under
75
the name of "white khutu," quite apart from the fact that
rhinoceros horn is not white. The possibility of walrus tusk
with which to a certain extent the rarer narwhal tooth is tied
up because both come from the same northern regions was
suggested by B. Laufer in the most extensive and learned
76
studies on the subject. Strong as his case seemed to be in
19 13 and 19 1 6, especially for the Chinese variety, the ku-tu-
77
hsi, it is now somewhat weakened by the information pro-
vided by the Hudud al-Alam that khutu horns were exported
from specific central Asian mountain countries, including
Tibet. It seems rather difficult to imagine, as Laufer did in
the case of the Kirghiz country, that the well-defined regions
mentioned in the Hudud al-Alam were only the intermediary
73 I.e., Wiedemann (164, 1916, p. 380), Reinhart (219, p. 184), Ruska (222,
p. 163). A. Zeki Validi Togan has lately expressed the same opinion (134,
p. 216).
74
164, 1916, p. 381.
75 This point was made by 164, 1913, p. 354.
76
164, 1913 and 1916.
77 In the writings of Dr. Laufer, ku-tu-hsi was transliterated as ku-tu-si.
exchange posts for the walrus tusks on their way from the
North to the Muslim world. Thus the newly gained informa-
tion makes the issue even more complex instead of clarifying
it. To animal geography and commer-
settle this question of
cial history, with the limited amount of information at hand,
seems still an impossible task.
Some observations can be supplied, however, which may
provide at least a partial solution of the problem. First, it
seems that one should not treat the ordinary colored khutu
and the white variety as one and the same material as was
done in earlier investigations. It is true that the two types of
horn have certain features in common and have thus, in the
mind of the medieval Muslim, been connected with each other,
even fused to a certain degree yet they were originally not ;
identical. In the case of the two types of khutu, as we have

shown above, the text of al-Biruni distinguishes between the
animals from which they came, the different material quali-
ties of the horns, the countries from which they were imported,
and the people who imported them. Considering all possible

solutions, it seems most likely that the white khutu, the fish
tooth from the northern sea, is either the tusk of the walrus
or that of the narwhal.
The identification of the colored khutu is a much more com-
plicated question. Our oldest and best authority, al-Biruni,
makes it clear that the material does not look at first sight
horn or teeth. He states twice that the khutu is "the bone
like
from the forehead of a bull," and he concludes only later on,
after the discussion of the fish tooth called white khutu, that
"judging from its appearance" this regular khutu is "like-
wise the main portion of a tooth or horn." Therefore
. . .
one has to visualize this khutu as something odd looking and

perhaps fragmentary. The mountainous regions mentioned
in the Hudiid al-Alam as the sources of supply seem to point
to a land animal. Thus the possibility of an identification with
mammoth bones, a suggestion first made by Wiedemann 7S
but
78
277, p. 354, footnote 2. Later on Wiedemann thought khutu to be rhinoceros
horn (164, 1916, p. 380). About mammoth teeth in the Muslim world, see
143, p. 18, and 177, p. 319.
later rejected by Laufer, 79 poses

itself again. In view of the
vagueness of the original sources and the confusion and wrong
identifications in the Middle Ages it seems, however, not very
fruitful to speculate about a solution of this riddle, especially
as it is of no immediate significance for this study.
Let us now turn back to our main thesis, namely, that one
variety of the khutu is the tusk of the walrus or that of the
narwhal. That we speak here in the same breath of two teeth
of such different form and character does not have to worry
us unduly, since the medieval zoologist was, as we have seen,
notoriously unconcerned about such matters. As long as the
animals or their horns had some identical basic features they
were easily fused, even in the mind of the scholar. The nar-
whal tusk was probably quite rare in comparison with that of
the walrus. Furthermore, it can be assumed that the teeth
came into the trade as broken or cut pieces, so that only the
traders and the artisans knew the difference.
The various arguments for our assumed identification of
the "white khutu" are individually perhaps too limited to be
conclusive, but when put together they all point in the same
direction.
There is first the Chinese ku-tu-hsi, which eminent oriental-
ists like Laufer, Pelliot, and Ferrand have identified with the
walrus and narwhal tusks, so that at least the nature of the
80
khutu's Chinese counterpart is clearly established. More
significant, however, is the evidence relating to the use of
fish teeth in the Near East.
Let us begin with the rarer narwhal tusk. There is first as
a most important piece of evidence the tusk in the treasury of
San Marco which reached Venice in 1488 (pi. 47, upper and
middle). 81 It must have come from the Muslim East because
79
164, 1913, pp. 329 and 354; 164, 1916, pp. 372-373 (though in 164, 1913,
p. 356, Laufer holds a confusion with mammoth ivory as a possibility).
80
164, 1913 and 1916; 98, vol. 2, p. 679. The first to connect khutu with
ku-tu-hsi was G. Jacob (143, pp. 82-83; 142, p. 9).
81 210-211. Schoenberger gives a detailed description
238, pp. 198-200, figs.
and history of the tooth. Its surface has been scraped smooth; the scrapings
were probably used for medicinal purposes.
naskhi inscriptions are engraved on its silver mountings. In

turning to literary references, one can point to Philip II of
Spain who is said to have received 12 narwhal tusks or
82
"alicorns" from the Sultan of Turkey. In 1623, according
to Pietro della Valle, European merchants still regarded Tur-
83
key as a likely place to sell such a tusk. Muslims, too, must
have participated in this trade because in the sixteenth cen-
tury the French traveler Andre Thevet reports that on an
island in the Red Sea and on the mainland nearby, he saw
men who artificially straightened the horns of elephants and
walruses and sold them as true alicorns. 84 Manipulations of
such a nature presuppose a knowledge of the real narwal tusk.
As to walrus ivory, historical and archeological evidence
shows that it appears fairly early in the Muslim world and
was then continuously used. Al-Biruni's statement that the
white khutu was brought by the Bulgars from the northern
85
seas and then made into knife handles is actually not the
earliest reference to it. A

'Awfi (which according
passage in
to J. Markwart goes back to a geographer of the first half
of the tenth century)tells us that in the Arctic Ocean "a fish
risesfrom whose teeth handles of knives and swords are

made." 86 Not quite so specific, but apparently still referring
to the walrus, is a passage in al-MaqdisI (end of the tenth cen-
82 quoting Immanuel Meteranus, Historia Belgica Nostrl, year

242, p. 112,
1599.
83 The passage is quoted verbatim in 166, pp. 367 f., footnote 3.
84
242, pp. 172-173.
85 This use is also attested for the ku-tu-hsi, 164, which dem-
1913, p. 318,
onstrates the Far Eastern usage for the first half of the twelfth century (see
also pp. 359 and 366) ; p. 358 gives a reference for the year 1320.
86 1 and 2), 310, and 312, where the passage is interpreted
177, pp. 265 (lines
as referring to the "Bartenwal" (i.e., Greenland whale) and its baleen, since
the author cannot imagine that one would have called the walrus a fish. This
paper has amply proven that zoological opinions of rare animals were very
vague in medieval Muslim society so that it does not seem astonishing that a
walrus should have been called a fish. Furthermore the baleen of the whale is
quite unsuitable for sword handles. Al-Biruni's statement that "white hafts of
knives are sawed from the fish tooth" also point clearly to maritime ivory.
As we shall presently see, Muslim swords with handles made of walrus tusks
are extant in large numbers. Hennig (121, p. 243) likewise thought this
passage referred to walruses and not to whales.
tury A.D.) according to which fish teeth together with other

northern products such as amber and sable and ermine pelts were
again imported by the Bulgars. 87 The early twelfth-century
Marvazi, in his chapter on the northern seas, speaks also of
88
a fish "whose tooth is used in setting knives." Unlike al-
Biruni, none of these sources employs the term khutu. There
is only one other author who provides a name. It is a different
designation, but details in this passage corroborate and supple-

ment al-Biruni's text and thus indicate that we are still deal-
ing with the same material. This refers to a definition of
chatuq J^, which in 1073 (466) Mahmud al-Kashghari in-
cluded in his Turkish-Arabic dictionary Dhvdn lughat al-Turk,
and which, on account of its importance, deserves to be quoted
in full: "Horn of a sea fish imported from China. It is (also)
said that it is the root of a tree. It is used for knife handles.
The presence of poison in food is put to test by it because
when broth or other dishes in the bowl are stirred with it the
food cooks without fire (if poison is present in it), or if the
horn is placed on a bowl it (the horn) sweats without
steam." 89 There seems little doubt that this refers to the tusk
of the walrus or the narwhal. 90 As in the case of al-Biruni
87 J ne 2 (the Arabic term used The passage

75> P- 3 2 5> li is I 0^-*"D' is
reproduced and translated in 143, p. 4.

88 quoted by Minorsky, Professor Minorsky was
183, chapter 15, ^[3, p. 83.
kind enough to refer me to this text.
s0 vol. 164, lines 9-12. In this edition is written with a jim,
3, p.
but Biockelmann who used this text transliterates it as chatuq, as if it had a

chim (58, p. 112). We are following Brockelmann's rendering of the word
in this investigation.
90 Mahmud's secondary explanation of chatuq as the root of a tree seems
to be based on the Chinese term ku-tu-hsi or at least on its first two syllables.
Although Laufer did not know Mahmud's passage, he came across the relation-
ship to a tree in some other context, so it seems best to quote him:
In the "peculiar Khitan writing of the word ku-tu (ku-tu-hsi is appar-
/f^/f^
ently of Khitan origin) the classifier 'tree' appears in either element The . . .
tradition of ku-tu being in structure or appearance like wood seems to have

originated in the Liao period (907-1125) and was perpetuated down to . . .
the age of the Mongols . . . Under the word ku the Dictionary of K'ang-hsi
cites the T'u king pen ts'ao of the Sung period to the effect that this word
refers to the trunk of a tree which is white like a bone, and hence receives its
name, and that the southerners make from it very fine utensils. Ku-tu does
(who had stated that the khutu was preserved in the trea-
suries of China), M ah mud links "the horn" with the Far
East, and he describes even more vividly the manner in which
the poison is detected, which is remarkably close to the
accounts which we find later in European literature. The
inclusion of chatuq in a Turkish dictionary confirms also the
information supplied by other authors namely, that the Turks ;
played a decisive role in the history of the khutu. As to the

difference in names, Brockelmann regarded chatuq as the
right form and assumed khutu and khartut to be corruptions. 91
not relate to any specific tree, but denotes the burls or knotty excrescences on
the trunks of the various trees which in diverse parts of the world, owing to
their fine veneer, are chosen with a predilection for carvings, particularly of
bowls. . . . The most
clever artists in burl-carved work known to me are the
Tibetans These bowls have two peculiar features in common with ku-tu-si:
. . .
many of them are white and yellow, and with their peculiar veins, offer some-
what ivory-like appearance; and some of them are believed by the Tibetans
to be capable of detecting poison" (164, 1916, pp. 359-360).
In view of the fact that Mahmud had stated that the chatuq came from
China it is thus not at all surprising to find this Far Eastern thought reflected
in its definition. The contradiction between the reference to a root as we find
it in the Muslim source and the knotty excrescence used in China can be
easily accounted for, since the two are not too different, at least in the eye of
the nonexpert. More important is that the appearance and the faculty of detect-
ing poison relate the wooden vessels to maritime ivory, and that Tibet, from
which the finest bowls came, was one of the regions which supplied the khutu to
the Muslim world (129, p. 92).
The identification of ku-tu-hsi with parts of a tree was, however, not only
a Far Eastern way of thinking. A certain wood found in Turkish regions
possessed so many of the features of the khutu that the two invited comparison
and, possibly, confusion in the minds of some people. Al-Biruni gives us the
following information about this wood:
"Theparticular use of the word khalanj is for a certain wood (showing . . .
figures and colored lines) of which, in the lands of the Turks, tables, beakers,
drinking vessels, and similar objects are made. Sometimes these patterns are
so fine that they resemble those of the khutu. When the markings are as fine
as this, knife and dagger handles are made of this wood which the Bulgars
import into Khwarizm or Khurasan." (134, p. 214, al-Biruni quotes Hamza
alTsfahani, who died between A.D. 961 and 971.) Here not only the close
relationship between the horn and the wood in appearance and special use
becomes apparent, but the text tells us also that both were imported by the
same people, the Bulgars. To this may be added that the khalanj wood occurred
also in the Kirghiz country, which likewise supplied large quantities of the
khutu horn which was there fashioned into knife handles (129, p. 96).
91
58, p. 112.
On the other hand, one should keep in mind that the Chinese
ku-tu-hsi points to an Islamic term more like khutu. Be that
as it may, in view of the comparative rarity of maritime ivory
it is good to recall the many variants for bishan ("rhinoceros
horn") 92 and to realize that it is not surprising that the tradi-
tion about the correct Arabic or Turkish name should have
been rather weak. It may very well be that khutu and chatuq
are but two readings of the same word, since the first two let-
ters are identical in Arabic and differ only in their diacritics,
while the final letters are close enough to make it possible for
their written forms to be mistaken for one another, especially
when the qaf has no diacritical points. 93
The use of maritime ivory for the making of knife and
dagger handles is also attested for later periods and by non-
Muslim Thus in 15 18, according to the Annals of
writers.
the Ming Dynasty, a Muslim ruler sent knives made of fish
teeth, together with horses and camels, as tribute to the Chinese
94
court. Several Western authors of the sixteenth and seven-
teenth centuries likewise mention that daggers with skillfully
fashioned handles of walrus tusk (often called fish teeth)
95
were very popular in Turkey at that time. The Jesuit Father
92 See above, p. 94.

93 The uncertainty of the correct spelling of the word is also demonstrated by
a single page in Yaqut's Geographical Dictionary which contains first the
word Jj-^l and then J ^>-\ (281, vol. 3, p. 447, lines 5 and 21). Both words
do not make sense and have thus been emended to stand for al-khutu, the
first change having been suggested by P. Pelliot (98, vol. 2, p. 679) and the
second by Wiistenfeld (281, vol. 5, p. 290). Jj-~-1
reflects the spelling J^v
as given in Mahmud al-Kashghari's dictionary, since only the diacritical points
of the first two letters are different.
94 quoting the Annals of the Ming Dynasty.
164, 1916, p. 355,
90
These passages are to be found in 164, 1916, pp. 361-365, quoting von
Herberstein (1517 and 1526), Belon (1553), and Avril (second half of the
seventeenth century). Laufer (164, 1916, pp. 365-366) points also to a passage
in Jahangir's memoirs which speaks of a Persian dagger with a fish-tooth hilt
spotted with black. This was sent to the Emperor as a gift of Shah 'Abbas of
Persia (144, vol. 2, p. 94). The curious piebald look of the "fish tooth" induced
the editors of the memoirs to assume that the material was probably tortoise
shell. There are however, fossil walrus teeth which have just this spotted
appearance. The memoirs go on to tell how the emperor tried to get another
piece like the fish tooth of the presented dagger, how it was finally found and
then made into two dagger hilts and a thumb stall (144, pp. 96-98).
Avril, for instance, states that "the Persians and Turks who
buy up the tusks of the walrus put a high value on them and
prefer a scimitar or a dagger haft of this precious ivory to a
handle of massy gold or silver ." 96 Although the sources . .
are usually specific in speaking of walrus tusks in connection

with these weapons, in one case the description could possibly
refer to a narwhal: "When the Turkish merchants buy the
knives they give them to the workers so that they put a point
on them, which is usually the tooth of the walrus of which
there are two sorts. One is straight, white and compact,
resembling the alicorn; . . . the other is curved like that of
a boar . .
As a final reference we should like to quote a passage from

Of the Russe Common W
ealth by Giles Fletcher, published
in London in 1 59 1 which seems to bind together all the vari-
,
ous aspects of the problems: "Besides these (commodities)

they have . . . the fishe tooth . . . which is used both among
themselves and the Persians and Bougharians, that fetcht it
from thence for beads, knives, and sword hafts of noblemen

and gentlemen, and for divers other uses. Some use the pow-
der of it against poison, as the unicornes home. The fish that
weareth it is called a morse, and is caught about Pechora.
These fishe teeth, some of them are almost two foote of
98
length, and weigh eleven or twelve pound apiece." Here
we first have the identification of "fish teeth" with the tusk
of the walrus or morse, and then the confirmation of its use
for knife handles and sword hafts; in addition the Bulgarians
are mentioned as intermediaries in the commerce; then comes
the important allusion to the antidotal property which we have
all along assumed for the fish tooth, owing to its connection
with the colored khutu which is now proved, and finally we
find the mouth of the river Pechora mentioned as a hunting
ground for the walruses; was the goal of many this river
trading missions, since the region between it and the Ural
mountains supplied the Muslim world in the Middle Ages
m 164, 1 91 6, p. 364.
07 According to Pierre Belon (quoted in 164, 1916, pp. 362-363).
ya
50, p. 13; also quoted in 166, p. 567; 242, p. 133.
with the most coveted furs and it is evident that it was also
the natural highway for the trade in walrus tusks.
In view ofall these references in Oriental and Western
sources seemed imperative to examine knives and swords

it
of various types from Muslim countries, especially Turkey and

Iran, to find out whether walrus (or narwhal) ivory was
used for their handles. About 50 swords, daggers, and knives
with ivory handles in the Walters Art Gallery, Baltimore, the
Metropolitan Museum, New York, the Peabody Museum,
Salem, and the United States National Museum, Washing-
ton, D. C, were therefore investigated. It soon became evi-
dent that most handles were fashioned from the tusk of the
walrus, whose tooth structure, with its characteristic osteoden-
tine core, makes it easy to identify." For instance, all the 17
swords, daggers, and knives from Turkey and Iran, dating
from the sixteenth to the nineteenth centuries, in the Walters
Art Gallery have handles made of the tusks of the walrus
(pi. 47, lower). There seems, therefore, little doubt that the
hilts and handles of many other swords and knives are made
1
99 Excellent analytical descriptions and illustrations of the various aspects

of the elephant, hippopotamus, walrus, and narwhal ivory are given by Schoen-
berger (238, pp. 171-174 with figs. 190-192). They clearly indicate that the
material of the handles is derived from the walrus.
It has been suggested that some of these handles might have been made
from the teeth of the sperm whale which lives in the warm parts of all oceans.
I am greatly indebted to Mrs. W. E. Schevill, of the Museum of Comparative
Zoology of Harvard University, for having been kind enough to make sections
of the tooth of this whale and of the tusk of the walrus so that this question
could be settled. They showed clearly that the dentine and osteodentine struc-
ture of the whale tooth is quite different from that in the material used for
the handles. Although the osteodentine section in the latter seemed often
unusually wide, there seems little doubt that the material comes from the
tusk of the walrus.
1
In the Henri Moser-Charlottenfels were many knives and
collection of
swords of various shapes with hilts made of walrus ivory. These arms came
from Iran, Afghanistan, Turkey, Bukhara, and Khiva, and dated from the
seventeenth till the nineteenth century. The collection contained also two
powder priming flasks of that material. The excellent plates of the catalog,
especially those printed in color, make it quite easy to recognize the walrus
ivory (189, pis. 8-10, 12, 17, and 40, in color 11, 18, and 19).
of the same material. 2 Since the literary evidence has been so

fully proved for the later centuries, it can readily be assumed
that the medieval armorer likewise used walrus tusks for his
swords and knives. So far no medieval Muslim knives, dag-
gers, or swords with ivory handles have come to my notice.
8
The Muslim swords and knives in our museums corroborate

still another point of the literary sources. Al-Biruni states
4
that the middle portion of a fish tooth, sawed into pieces, is
distributed among the various hafts so that every piece is
shared by them. These sections are said to indicate that the

haft is made of the tooth and not of ivory or from chips from
the edges, because they display various designs which have the
appearance of "wriggling." Al-Ghaffari deals with the same
aspect of the material (which according to him is yellow to
red) and speaks of damascening designs. These authors are
obviously referring to the inner part of the walrus tusk with
its characteristic osteodentine core showing a densely veined
pattern in a slightly browner or more reddish color. On ex-
amination of the swords and knives it at once becomes evident
that the sections of the tooth are so applied that the finely
patterned osteodentine part shows prominently. It is there-
fore easy just as al-Biruni said to recognize a haft made
of walrus tooth and to distinguish its material from elephant
ivory, which lacks this particular design.
2 It is called ^ft>^ dandan-i mahi or ^^bb shir-mahl or shir-i

mahi in Persian.
Stone (251, p. 561, s.v. shirmani [sic]), referring to the use of walrus tusks
for hilts in the Near East, states that the reason for this custom is that this
material is less likely to split than elephant ivory. Since Stone does not give
his source it was not possible to check on this information. Another explana-
tion is given by Father Avril who states that the tusk has the property of
stanching blood (164, 1916, p. 364).
3 K. E. von Baer has shown that the use of maritime ivory for sword hilts
is referred to as early as thefirst half of the third century A.D. At that time
Solinus reported that the inhabitants of ancient Ireland used the teeth of mari-
time animals for that purpose (34, pp. 126-130; see also 164, 1913, pp. 333-
334).
4 Wiedemann
(277, p. 354), in his translation, speaks of the "mittlere Teil
(des Zahnes)"; he obviously refers to the core of the tooth which alone shows
the characteristic pattern.

The Middle Ages walrus ivory was avail-

fact that in the
able in the Muslim East
in large enough quantities to be used
in the manufacture of such ordinary implements as knives and
daggers raises the question of how this material reached the
Near East. Al-Biruni tells us only that it was the (Volga-)
Bulgars 5 who imported the fish tooth from the northern seas.
From other sources we learn more about their transactions
for instance, that their merchants brought unpolished swords
from Azerbaijan to the heathen Wisu (or Isu), one of the
northernmost people known to the Muslims, 6 to be traded
there for beaver pelts and other furs. 7 The Wisu, in turn,
bartered the swords for sable furs in the lands farther north, in
the regions along the Arctic Ocean where the blades were used
8
as harpoons in the hunting of whales. It can be assumed
that the walrus ivory reached the Muslim world via the fur-
trade routes. 9
Trade in walrus ivory on the north coast of Russia (where
5
The capital of the Bulgars was about 72 miles (115 kilometers) south of
Kazan (see 129, p. 461, and map 12 on p. 435; 35, vol. 1, p. 788).
6 Usually identified with the Finnish Ves' foot-
(129, p. 309; 134, p. 55,
note 3). Hennig thought that they lived in the Perm region, the center of
which was Cherdyn (121, pp. 254-255). Zeki Validi Togan rejects this identi-
fication and proposes Kotlas (134, p. 172). According to Ibn Fadlan the Wisu
lived a distance of 3 months from the Bulgars (134, p. 55; 177, p. 282). This
applied to the water route during the summer months; in winter it took only
20 days to cover this distance with sleighs (134, p. 171).
7
177, P- 300; 134, P- 67.
8
(143, pp. 76-77; 177, p. 300; 99, p. 24). The people to the north of the
Wisu probably lived even beyond the Yura or Yughra (177, p. 301), identified
with the Ugrian Ostiaks and Voguls (129, p. 309).
The location of the various far-northern tribes referred to by Muslim
writers is still a matter of conjecture. The same applies to the possible routes
which the traders used during the summer and winter expeditions (Togan,
I 34> PP- 171-173, objects to Hennig, 121, p. 251). But the existence of the
trade is well established, as are many details of the way in which it
functioned.
9 Another proof of the existence of this fur trade is the praise which Ibn Sa'Id
bestowed about 1260 on the furs of the polar bears which were brought to
Egypt (121, p. 249). In the fourteenth century Arab traders occasionally man-
aged to reach the Far North, since we have reports about their trips to the
northern Ural Mountains (121, pp. 259-260) and they even penetrated as far
as the regions of the fur hunters along the shores of the Arctic Ocean (121,
p. 263).
both the walrus and the narwhal live) 10 is even older than
the earliest references to it in Muslim literature. In the
second half of the ninth century A.D. the Norseman Ohthere
sailed around the North Cape to "Biarmia" (the region from
the White Sea to the Ural) to obtain "horshvael which have
11
in their teeth bones of great price and excellencie." This
may have started (or stimulated) commercial activities in this
commodity with Europe.
As to the Muslim trade with the tribes of the Far North
we need not rely only on literary sources, since it is also at-
tested by archeological finds. Two silver vessels with Kufic
inscriptions of the eleventh to twelfth century were found
below the Arctic Circle in northwest Siberia, one in the Dis-
trict of Berezov, and the other in a fortified camp on the
12
river Sosva. In addition there have come to light in the same
general region three other silver vessels whose more barbaric
style indicated that they came from Muslim borderlands. 13
All these objects, like the swords from Azerbaijan, must
have been exchanged for the goods of the northern regions,
14
that is, the much-coveted furs and the maritime ivory.
While literary references and actual objects make it cer-
tain that walrus tusks and, to a lesser degree, also narwhal
teeth were known in the Islamic East and that they are obvi-
10 One species of the walrus, Odobenus rosmarus L., lives in the Arctic Ocean
in an area stretching from the mouth of the Yenisei, around Novaya Zemlya,
Spitsbergen, and Greenland as far as Hudson Bay. It has a shorter and more
curved tusk than the North Pacific species, Odobenus obesus 111., which lives
along the coasts of northeast Asia and northwest America (53, vol. 12, p. 629).
The narwhal, Monodon monoceros, lives usually between the 70th and 80th
degrees northern latitude and is to be found among other places around
Novaya Zemlya and the waters north of Siberia (53, vol. 12, p. 476).
11
34, PP- 7-8 and 116-125; 164, 1913, pp. 337-338.
12
223, pi. 82, No. 147, and pi. 83, No. 148, which date from the twelfth
century (sixth century) (262, pp. 406-407). The sites are indicated on map 1
in Smirnoff's introduction (223). Another silver vessel of slightly earlier date
was found farther south in northwest Siberia, near Surgut (223, pi. 80, No. 145).
13 No. 92; pi. 85, No. 155; pi. 87, No. 156.
223, pi. 58,
14 Walrus ivory also reached Europe, as shown, for instance, by two twelfth-
century German walrus carvings in the Walters Art Gallery, one a chess piece
(113, vol. 4, No. 284), the other part of a larger composition, probably a
corner of a portable altar (221, fig. on p. 243).
ously the "white khutu" or fish teeth of the ancient sources, it
still has to be demonstrated that the medieval Muslim actu-

ally confused the khutu horn with that of the rhinoceros. This
final link is necessary to explain how in Islam the antidotal
virtue of the khutu became suddenly attached to the horn of
the karkadann.
There is literary proof that this identification of the khutu
with the rhinoceros horn happened at least as early as the be-
ginning of the twelfth century. About 1120 the Persian geog-
rapher Marvazi states in two places that the khutu is the
horn of the rhinoceros and, vice versa, that the rhinoceros
horn is called khutu. To make his assertion even more specific
he mentions, too, that the Chinese make high-priced girdles
from this khutu; 15 this was, as we have seen, universally and
specifically stated about the horn of the karkadann.
Marvazi does not stand alone with his identification. About
200 years later al-Akfani states that the khartut o y J is also
called khutu. Khartut in the form of khartit oJ J>- or khartit '
JaJ* J>- is Egyptian designation for a rhinoceros. 16

still the
Thus khutu horn and rhinoceros horn appear again as inter-
changeable terms. 17
15 paragraph 15) and

183, Arabic pp. 5 and 11, translation p. 17 (ch. 8, ;
p. 23 (ch. 8, paragraph 27).

The mistake in the Muslim identification is clearly demonstrated by a Chinese
source which is approximately contemporary with Marvazi. Flung Hao states,
about 1 143, in the Sung mo chi oven: The ku-tu-hsi (the Chinese eqivalent of
the khutu) is so rare that "among numerous pieces of rhinoceros horn there
is not one (of this kind). It has never been worked into girdles (as in the
case with rhinoceros horn)" (164, 1913, 359; about the date see 164, 1913,
p.
p. 366). Marvazi's identification shows that Laufer's statement "No Arabic

author has ever used the word khutu with reference to the rhinoceros" (164,
1916, p. 381, footnote 3) can no longer be maintained.
16
164, 1916, p. 380, quoting Wiedemann; see also 88, vol. 1, p. 363 (JaJ? j> ) ;
98, vol. 2, pp. 679-680; 92, p. 182 (C~> j>) ; 53, vol. 12, pp. 605 and 607. It
will be recalled that Brockelmann assumed both khartut and khutu to be

garbled versions of chatuq (see p. 123).
17 This opinion is exactly the opposite of that of Laufer who states that "if
anything is quite certain, it is that rhinoceros horn is not understood by chutww
{khutu) (164, 1913, p. 354). Laufer assumed also that "the Arabs have . . .
merely transferred to the walrus-tusks certain popular beliefs entertained

regarding rhinoceros horn," while my deduction is that the influence ran in
Once the equation between the two materials was made, it

was natural that the special quality of detecting poison was
sooner or later transferred from the khutu to the karkadann.
There is as yet no indication of this transfer in Marvazi's
long chapter on the rhinoceros, perhaps because the identifica-
tion was made only shortly before this writer's time. To our
knowledge the first reference to poison appears in the report
of al-Qazwini, where the new virtue of the rhinoceros horn is
significantly mentioned after it has been stated that it is good
for knife handles. 18 Al-Qazwini (and after him al-Mustawfi)
seem to have regarded it as an antidote since he stated that
.when it is near poisoned food or drink it breaks the
.
strength of the poison." There is not yet any reference to the

poison-detecting properties which distinguish the khutu.
The final stage in the development was reached when the
khutu's agitation or sweating in the presence of poison was
attributed to the horn of the rhinoceros. This point was
reached about 1446 (850) in the Kharidat al-'ajaib wa-faridat
al-gharalb the cosmography of Ibn al-Wardi. 19 In this work
y
one learns that "one utilizes the horn of the rhinoceros by

using it for the handles of the royal table knives. When
poisoned dishes are placed before the king the handles be-
come moist and agitated." 20 Here then are all the elements
the opposite direction.Laufer's research has, however, immeasurably enriched
our knowledge and his conclusions, which no longer seem to tally with histori-
cal data, are due only to the fact that he did not have enough Islamic source
material at his disposal.
18 Besides the poison-detecting qualities there is yet another medicinal prop-
erty which was transferred from the khutu to the karkadann. Al-Akfani
reports "that it has been established by experience that together with the vapors
of perfume, the khutu has an excellent effect in the case of hemorrhoids." This
is precisely the effect ascribed to it in India where for this reason rhinoceros
horn is used in the form of plates for food, flnger-and-toe-rings, cane handles,
and so on (219, p. 184), although the earlier accounts of the karkadann never
mention this particular use of the horn.
19 Ibn al-Wardi is not an original writer: his book is only a transcription,
often word for word, of the Jami' al-funiin of the Hanbalite geographer Najm
al-din Ahmad b. Hamdan al-Harrani, who was in Egypt in 1332 (732) (see
98, vol. 2 p. 412; 57, vol. 1, pp. 130-131 and suppl., vol. 2, pp. 161-163). Ibn
al-Wardi's passage might therefore go back to al-Harrani.
20 At the beginning of his statement about the rhinoceros Ibn al-Wardi quotes
of the myth in the proper setting: the problem of detecting

poison on the royal table, the knife handles made of a special
horn, and, finally, the revealing symptoms of sweating and
agitation. While in the case of al-Qazwml one is still left in
the dark about a possible connection between knife handles

and the poison-detecting quality of the horn, their mutual re-
lationship is now established, and the entire legend is presented
in a logical and well-integrated fashion. The whole reads now
like an account of the alicorn's magic power, written in the

West, except that Ibn al-Wardi concludes his report with the
usual Muslim story of the Chinese girdles made of the horn
and their great monetary value.
From this time on the horn of the rhinoceros seems to have
been endowed with the mythical antidotal property, though
the decline of great writers in the Muslim world makes it
more difficult to establish evidence for this; but even so there

are occasional references especially in Western sources. For
instance, James Lancaster, while lying in the Straight of
Malacca in 1592, tried to barter for the horn, "which is
highly esteemed of all the Moores in these parts as a most
21
soveraigne remedie against poyson." Even in our times
vessels made of the horn are said in Egypt, India, and Tur-
key to possess the antidotal quality. 22
al-Jaihani as his source for the information that the karkadann in the shape
of a donkey (see pi. 20) is to be found in Sumatra. Since al-Jaihani's work,
written between 892 and 907 (279 and 295), is now lost, it is not possible to
check on whether any of Ibn al-Wardi's additional remarks go back to this
early source. It can, however, be assumed that since antitoxic reactions of
knives are absent in ail earlier reports, al-Jaihani can hardly have given this
information.
21
242, p. 218.
22 For Egypt see the report of M. Meyerhof in Cairo (98, vol. 2, p. 680),
and for India those of Reinhart in Delhi (219, p. 184) and of Shebbeare quoted
in 118, p. 377. Andersson reported in the middle of the last century that cups
made of horns of reddish tints were also "esteemed" in Turkey on account of
their poison-detecting quality (25, p. 309). The same author gives in addition
two other accounts of the powers of the horn which so vividly describe the
traditional Muslim beliefs that they are herewith quoted:
" 'The horns of the rhinoceros,' says Thunberg (apparently Karl Peter Thun-
berg, 1743-1828, aSwedish naturalist), 'were kept by some people both in town
and country, not only as rarities, but also as useful in diseases and for the
After having followed the later history of the horns of the

khutu and karkadann and their relationship to poison, it might
be appropriate to deviate for a moment and ask when and
where this whole development started. Since both al-Biruni
and Marvazi have pointed to a connection with China, one
naturally looks to Far Eastern authors for further elucida-
tion of this question; but here we run into difficulties. Al-
though Laufer investigated the history of the walrus and
narwhal tusks in China, where they had been known since the
T'ang period, he could not find anything about the ku-tu-hsi's
connection with poison before the middle of the thirteenth
century, that is to say nearly two centuries after the Muslim
authors had spoken about it. An earlier Chinese account tells
us only that it was a priceless material, veined like ivory and
of yellow color, that it came from the Khitan country in the
far north of China and that it was used for sword hilts and
knife handles.
-'
1 3
When, about 1259, the ku-tu-hsi is finally
mentioned in connection with poison it is listed as a product

of the Muhammadan countries of the West or of central Asia.
Its outstanding quality is not the detection of poison due to
a specific reaction, as al-Biruni and al-Kashghan had indicated,
purpose of detecting poison . . . The fine shavings of the horn taken internally
were supposed to cure convulsions and spasms in children. With respect to
the latter itwas generally believed that goblets made of these horns in a
turner's lathe would discover a poisonous draught that was put into them by
making the liquor ferment till it ran quite out of the goblet. Such horns as were
taken from a rhinoceros calf were said to be the best and the most depended
upon.'
" 'The horn of the rhinoceros,' Kolbentells us, 'will not endure the touch of
poison. have often been a witness to this. Many people of fashion of the
I
Cape have cups turned out of the rhinoceros horn. Some have them set in
silver, and some in gold. If wine is poured into one of these cups, it immediately
rises and bubbles up as if it were boiling; and if there be poison in it, the
cup immediately splits. If poison be put by itself into one of those cups, the
cup, in an instant, flies to pieces. Though this matter is known to thousands
of persons, yet some writers have affirmed that the rhinoceros horn has no such
virtue. The chips made in turning one of those cups are even carefully saved,
and returned to the owner of the cup; being esteemed of great benefit in
convulsions, faintings, and many other illnesses.' " (25, p. 309.)
23The passage occurs in a treatise which was written pursuant to a mission
which took place from 1129 till 1143 (164, 1913, pp. 318-320).
10
24
but its neutralizing power. In 1366 another author inter-
preted this further by stating that "ku-tu-hsi is poisonous by-
nature and can thus counteract all poisons, for poison is
treated with poison." He arrives at this opinion because, as

he points out, one of the Chinese transliterations of the Khitan
word ku-tu-hsi is written with three characters meaning "black
25
magic, poison, and rhinoceros horn." The latter was re-
garded as an effective antidote and it is therefore natural that
this important quality of the horn was transferred to the ku-tu-
hsi and that they were eventually thought to be identical (in
26
1387). Even in the latest reports there is no word about
the special poison-revealing reaction which was attributed to
the khutu or the chatuq. We thus have either to assume that
there was an earlier connection between the ku-tu-hsi and the
horn of the rhinoceros than Laufer had been able to trace, or
we have to doubt the particular statements made by al-Biruni
and al-Kashghaii. If the second alternative should be cor-
rect
and it seems quite possible it may very well be that
the Turkish promoters of the maritime ivory used an alleged
Chinese reputation to enhance the value of their merchandise.
Let us now return to the iconographic aspect of our study.
Our thesis that the walrus and narwhal varieties of the khutu
became fused with the horn of the rhinoceros seems to be
borne out by many representations of the karkadann. Varied
as the body structure of the animal in the Islamic representa-
tions is, since the time of the Aleppo plaque of 1 1 68 and the
Mosul jar in the Berlin Museum, the majority show one iden-
tical feature : the long, straight, tapering horn. Even illustra-
tions of various iconographic types in al-Qazwini manuscripts

sometimes show this form in spite of the fact that the text
states not only that the horn is bent but goes so far as to
27
describe its curvature. The long, straight shape seems to
24 It occurs in the report of Chang Te, whom the Mongol emperor Mangu
sent in 1259 as a courier to his younger brother Hulagu, the founder of the
Tl-khan dynasty of Persia, who resided in Tabriz (164, 1913, p. 320).
25
164, 1913, p. 322.
26
164, 1913, p. 325.
27 "In it is curve convex towards the animal's front and concave towards
a
its back." This, by the way, is apparently the passage which has been mis-
have been regarded as so significant that it was kept even after

the rhinoceros had been recognized as possessing a curved
horn which the painters learned to place correctly on the nose
of the animal. Thus the miniature of the karkadann in the
Princeton al-Qazwini manuscript shows the fairly realistic
representation of a narwhal tusk, with its spiral grooves, in
combination with an ordinary rhinoceros horn (pi. 15, right).
There is, therefore, a good possibility that all the other long,
straight horns could be interpreted as tusks of the narwhal;
the slightly curved ones, like that on the head of the karka-
dann in the Sarre manuscript (pi. 13, lower), can be imagined
to have been inspired by walrus tusks even though the text
of al-Qazwini, which called for a bent horn, might have been
a contributory factor for this particular shape.
At this point in the investigation we are also justified in
assuming that the horns on the heads of two animals decorat-
ing the plaque from the Aleppo madrasa (pi. 6) possibly re-
flect a narwhal tusk. All the characteristic features, such as
the long, straight, and tapering form and the helical grooves
are unmistakably reproduced. Furthermore some of the later
representations of the karkadann (pis. 2, upper; 7; and 19,
upper) show vertical markings which may indicate that the
artists still knew that certain designs on the horn were neces-
sary, although they had forgotten the real ones and substi-
tuted more fanciful patterns. Thus the representation on the
Aleppo plaque might indeed be one of the earliest instances
in which the ideas of the terrestrial karkadann and the mari-
time khutu are fused. The only other example in an early
style so far found is the enigmatic representation on an His-
pano-Moresque ivory pyxis 28 which was formerly in the col-
lection of the Comtesse de Behague and is now in that of the
translated in Western books on the unicorn as follows: ". . . with raised

striae outside and a hollow within" (242, p. 265).
28 1 know this ivory box only from reproductions. It is discussed in 100,
vol. 1, pp. 75-76, and illustrated on plates 28 and 129. Ferrandis thinks that
the unicorns are fighting each other. Two pyxides, dated 968 (357) and 970
(359) (JOO, pis. 19-25), are said to be closely related. It is, however, quite
impossible that the Ganay pyxis is later than these parallels; it may even be
modern.
Marquis de Ganay (pi. 5, right). Here, too, each of the

paired unicorns proudly carries a long, tapering horn which
seems to be marked with helical grooves. These equine long-
bearded animals are, however, unique. All other unicorns on
Spanish ivory boxes of the same period, have a winged body
and carry a horn ending in a hook-shaped curve which, in spite
of the helical grooves, can have no connection with the nar-
29
whal tusk (pi. 5, left). The unicorns on the Ganay pyxis
have also no close parallels among the representations of the
species in the art of the Muslim East of the twelfth, thir-
teenth, and fourteenth centuries. On the other hand, they are
much Western types which have appeared in Christian
like the
30
art since the late twelfth century. In view of the unusual
character of these animals it seems wise not to draw any con-
clusions before other examples have been found or pertinent
texts have been discovered.
After having finished this survey we can now come back
to the final question which we raised at the beginning of this
chapter, namely, whether it was the karkadann which insti-
gated the medieval myth of the antidotal property of the ali-
corn. The complexity of the subject makes it impossible to

give anything but a tentative answer. With this limitation in
mind one can state that it could very well have been the horn
of the karkadann endowed with the new power it received
through its fusion with the khutu which was the source for
the medieval myth. The dates of the first reference to the
antidotal virtue of the khutu (in al-Biruni) and its first iden-
tification with the horn of the rhinoceros (in Marvazi) make
31
this quite possible. Only if the sweating of the alicorn should
prove to be an early and essential feature in the detection of
29 See also 100, vol. pis. 42, 48, 50, 56, and
1, 59.
30 196-197 and 207; see also 196, and
238, figs. pis. 2-4, 6, 7; 195, Nos. 73-78
and pi. 6; 94, No. 290 and illustration.
31 There would be thus three interchanges between rhinoceros horn and
maritime ivory. In China the antidotal virtue of the rhinoceros horn was
transferred to the ku-tu-hsi (164, 1913, p. 353), while in Islam the lore of the
khutu created in turn the antidotal myth of the rhinoceros horn. Finally in the
West it is now again the rhinoceros horn which seems to have endowed the ali-
corn with antidotal power.

the poison in the West (which it does not seem to be) would
the source of the influence be the khutu itself, because only in
the first half of the fourteenth century does the horn of the
karkadann allegedly show this peculiar reaction.
The possibility of Muslim influence on Western representa-

tions cannot yet be fully gauged. It is, first of all, tied up
with the unicorns on the atypical and problematic pyxis in the

Marquis de Ganay's collection. This document startles us on
account of the early manifestation of the type, just as it is
provocative in its historical implication. Nevertheless, its
uniqueness in Hispano-Moresque and seemingly even in Mus-

lim art does not admit any conclusions at this time. The second
earliest karkadann, with a straight, tapering horn and spiral
grooves, the one on the plaque of the Aleppo madrasa of
1 1 68 (564) ,
probably antedates by some years the earliest pic-
ture of the unicorn with a narwhal tusk in European besti-

32
aries. The difference in time is, however, very small and
the Muslim decoration, furthermore, shows only an animal
head and not the whole quadruped. Whether the third earli-
est Muslim piece, the Mosul jar fragment in Berlin, has a
chronological precedence over Western representations is al-
ready questionable. Its exact date in the late twelfth or, more
likely, in the thirteenth century, is difficult to determine, be-
cause its style is archaistic. Also, the long horns of its uni-
corns do not show the characteristic grooves. Thus, owing

to the scarcity and the problematic character of the examples
now available, the case for a priority of the Muslim unicorn
cannot, for the time being, be postulated, although it is not
impossible that in point of time the Near East might have a
slight edge over Europe. Only when we know more of early
Muslim representations of the karkadann, especially those on
easily transportable (and exportable) works of art, can we
be more precise about the possibility of a transmission from
East to West. For the time being it seems fair to assume
82 The date when in the West narwhal teeth were first thought to be the
horns of unicorns and represented as such has been put at about 1200 (238,
PP- I94-I95)-
:
that in medieval Europe, other and probably better circum-

stances must have existed in which the iconography of the
unicorn bearing a narwhal tusk could have evolved.
SUPPLEMENTARY NOTE
The account of the khutu in al-Biruni's Mineralogy, the
Kitdb al-jamdhir fi ma'rifat al-jawdhir, published in Hyderabad
in 1355 H. (see above, p. 114, footnote 57), contains the
following data
The khutu is a much-desired animal substance especially
treasured by the Chinese and the Turks of the East. Like the
bezoar stone, it is supposed to sweat when it is near poison,
a belief confirmed by messengers from the Qitay Khan, whom
al-Biruni had asked about it.
The exact nature of the substance is so uncertain that al-
Biruni gives several identifications and explanations. Some
statements are accompanied by comments, while others
critical
reveal clearly how confused the problem remained to him. It

is not surprising, therefore, when he finally states that to know
what the khutu actually is appears hopeless.
As in al-Khazini's version, it is first assumed to be a bone
from the forehead of an ox living in the land of the Kirghiz.
Al-Biruni gives little credence to this theory, though it is sup-
ported by older, unnamed books. According to his own ob-
servations, the khutu is too thick to be from such a bone,
especially since Turkish oxen are of small build. He
the
prefers to think that "it would fit the horn much better."
If this should be so, he would attribute it to the antelopes
of the Kirghiz. Another theory designates it the forehead
of a water karkadann (Û j-xS^S"), called water elephant
(Û Jus), while yet another characterizes it as a bone from
the brow or the horn of a very large bird. According to the
last assumption (which is the one to provoke the author's
final despair), the bird falls down on some island and the
bone is taken after the flesh has decomposed. Further details

in the ensuing account of the bird, such as its tremendous size
(which blots out the sun), its veneration as a divinity, its
habitat in deserts behind the sea between China and Ethiopia,

and finally, its feeding on elephants, make it clear that the
myth refers to the rukh.
The khutu's all-important pattern is discussed after al-
Biruni has mentioned the water karkadann as a possible

source of this substance. He likens its damascened design to
the core of a fish tooth which the Bulgarians import into
Khwarizm from Northern Sea. He then mentions that
the
the fish tooth is a little more than a cubit long and that its
core, running through its whole length, is known as its sub-
stance or jewel ( ^J] j^y? ). It is not quite clear whether in
the next paragraph he still speaks of the fish tooth (obviously
the walrus tusk) or the khutu, though it seems as if al-Biruni
again had the former in mind. It tells how a Khwarizmian
came across specimens whose enamel surrounding the core
was brilliantly white. He carved knife and dagger handles
from its core, which was white with an admixture of yellow
and cucumber split lengthwise. He brought
like the core of a
it to Mecca khutu and sold it to the Egyptians for
as white
a high price. Here our author is apparently referring to a
dishonest commercial transaction.
Al-Biruni stresses the khutu's connection with water in one
more instance. When carved pieces fall into fire, a fishlike
odor rises, and just as fumigating with fish bones is said to

help hemorrhoids, so does khutu smoke.
The description of colors is fairly close to al-Khazini's ver-
sion. Its beginning and end, however, are different. The
former is significant, since it could just as well apply to the
walrus tusk: "The best is the crooked one, which is colored
between yellow and red." The price of such crooked speci-
mens is 100 dinars (gold coins) for a hundred dirham's weight
(ca. 314.8 grams), while the poorest quality of the tooth costs
only one dinar with no regard to weight. The largest piece
seen by al-Biruni weighed 150 dirharns (ca. 472.2 grams),
and its value was 200 dinars.
Finally, al-Biruni reports two peculiar uses for the khutu
not otherwise mentioned in the literature. The Amir Abu
Ja'far b. Banu had a large chestlike box with long, broad, and
thick khutu plates, while the Amir Yamin al-dawla owned an
inkwell which was a source of blessing for him but brought
ill-luck to others.
The editor of the Kitdb al-jamdhir ft ma'rifat al-jawdhir

adds to his text (45 a p. 208, footnote 2) the passage on the
>
khutu in the same author's Pharmacology, the Kitdb al-

saidana. This is only a shortened version of the account in
the Kitdb al-jamdhir. In parts it gives the impression of being
slightly garbled, as is evident from a passage like: "the khutu
. . . is the forehead of a karkadann, that is the water
elephant."
Al-Biruni seems to imply that he has actually seen khutu
pieces. It is therefore significant that he distinguishes between
the khutu and the fish tooth, i.e., the walrus tusk. The selling
of a fish tooth to Egyptians as white khutu and the long,
broad, and thick plates of Abu Ja'far's box speak also for the
existence of two different materials. There are close similari-
ties, however. The khutu's core has designs and colors very
much like that of the walrus tusk, while the piece carved by
the Khwarizmian looked like white khutu (a substance which,
contrary to my assumption based on Wiedemann's translation
of al-Kazini's text, does not exist as a separate category).
The issue then is to find two animals which have "teeth" of
marked similarities. In this respect, it will be recalled that
I have come across a number of swords the handles of which
were unusually wide for walrus tusks, although they had its
osteodentine pattern.
Unfortunately, I cannot offer any satisfactory solution of
the problem. It seems unlikely that al-Biruni is speaking of
the tooth of the sperm whale which otherwise could figura-
tively be called a water karkadann or water elephant (see
p. 126, footnote 99). The narwhal has no core with a design.

On the other hand, it is not likely that Egyptians would have
paid a high price for hippopotamus teeth, which must have
been fairly common in their country. Furthermore, this animal
was known as "water horse" (*Ul ov), and the structure of
its tooth differs greatly from that of the walrus. In case there
is no other tooth like that of the walrus -nor a horn re-
sembling it the only remaining possibility would be that al-
Biruni makes a distinction between two types of walrus teeth,

perhaps teeth of different sizes or in different stages of preser-
vation; or we would have to assume that the cause of the
whole confusion is of semantic nature. It could be conjectured
that at a certain time the walrus was usually called a fish, but
sometimes also a water karkadann, without its being under-
stood that the two names referred to the same animal. Ac-
cepting this, it is only natural that the products of these
alleged two animals look very much alike, since they come
from one type animal. However, even if we have to accept
that, originally, there were two different products from two
different animals, it is clear that such rare and mysterious ma-
terials as khutu and walrus tusk could easily be mistaken for
each other, as they seem to have been in Egypt. This makes
it possible to understand how features connected with the
original khutu could have been transferred to the walrus tusk.
On the other hand, al-Biruni's identification of the khutu with
an ox or a water karkadann was the starting point for the
confusion with the terrestrial karkadann which was supposed
to have had a bovine form. If the text of the Pharmacology
is not garbled, this confusion occurred already in al-Biruni's
own writing.
Why our author called the tusk a forehead bone remains
another puzzle, especially since he himself preferred to call
the khutu a horn. Possibly he became prejudiced by the intel-
ligence given to him by the men of the Qitay Khan who
designated it as such or by the reports connecting it with the
rukh.
The passage in al-Biruni shows that the bird theory explain-
ing the nature of the khutu actually has two sources: the
buceros (see p. 116) and the rukh. The association of the
khutu with the rukh has its parallel in the relation of the
karkadann and the rukh (see p. 33). There is, therefore, one
further reason why khutu and karkadann could be confused
with each other. It is also noteworthy that al-Biruni is the

first to identify the khutu with the fabulous bird and that al-
Akfani repeats in his account only what he had read. We
likewise find that the same fourteenth-century author copied,
besides other facts, al-Biruni's original statements of the
khutu's popularity with the Turks, its range of colors, and its
usefulness in the case of hemorrhoids. In al-Akfani's time the

price of khutu seems to have dropped, however.
THE KARKADANN AS A SCIENTIFIC AND
ARTISTIC PROBLEM
The preceding sections dealt first with the figural represen-
tations of the rhinoceros in the Muslim world and then with
the lore of various unicorns in relation tomyths in other civi-
lizations. Extensive as our account may have seemed, its enu-
meration and reconstruction of facts do not tell the whole
story, since they have been presented in a historical vacuum.
It is necessary to survey the whole development and to con-
nect it wherever possible with the framework of intellectual
growth in the Muslim world. After all, the myth of a strange
animal represents a challenge to human resources and imagi-
nation and the approach to the problem reveals clearly the
power of observation in explorers and the critical faculties
in writers in each given period. In the case of the karkadann
writers set the pace for artists because the former were the
first to deal with the problem and it is on them, as we have
seen, that the artists heavily leaned. It is, therefore, with

them that we have to start the final part of this investigation.
When an author was confronted with the accounts of sailors
and problem was to decide, and this far
travelers, his first
away from the habitat of the karkadann, whether this strange
animal was just a human fabrication or a reality. As stated
in al-Jahiz' Kitdb al-hayazvdn, this was still an issue as late as
the ninth century. The attitude of al-Jahiz with regard to the
doubted existence of the beast was based on literary learning;
since it is mentioned in the Bible and in Aristotle, there is no
question in hismind that its reality is assured and so he pre-
sents whatever information he has at his disposal. 1 There is,
however, a certain amount of critical attitude in him and in
the best writers of the ninth, tenth, and eleventh centuries.
Thus al-Jahiz uses logical reasoning when he tries to challenge
the story that the young karkadann puts its head out of its
1
145, vol. 7, p. 40.
143
mother's womb and eats leaves when passing trees. His argu-
ment runs like this: "What it has eaten must necessarily be
excreted. If the young karkadann continued eating and did
not excrete, it would be remarkable; if it excreted in the
womb of its mother, it would be even more remarkable." 2

His conclusion is that he does not think that this is entirely
impossible, but he is inclined not to believe it. For al-Mas'udi
this story is and he uses a scientific
incredible method to check
it. He interviews men who have traveled in India as to the
mode of pregnancy of the rhinoceros and their answer dis-
3
credits the whole yarn. Al-Biruni, too, had a scientific atti-
tude when he clearly distinguished between his own observa-
tions and what he has been told by others. It is true, the
reports of his informants are full of fantastic details which
al-Biruni does not challenge, perhaps because he lacked proper
means of checking them, but he is at least fully reliable as to
what he has seen. This was no mean achievement considering
the preconceived ideas about the subject and the uncritical ac-
ceptance of book learning. One has only to compare his ob-
servations with those of Ibn Battuta, whose preconceived no-
tions of the rhinoceros obscured his perception of the animal
on the several occasions that he saw it.
Al-Birunl's independent attitude is likewise shown in his

account of the mysterious khutu. He quotes in the first place
an opinion which is founded on data provided by foreign in-
formants thought to be familiar with the subject. Yet, in

spite of the fact that their information is corrobated by liter-
ary sources, he questions this lore from his own observations,
gives the reasons for his doubts, and with all due caution con-
cludes by providing his own theory. In the end, he does not
hesitate to throw doubt on all his given information because
he sees the hopelessness of reconciling the various contradic-
tory, if not outright confused, sources.
Al-Nadim, another figure of the early period, is the only
person who uses the literary form of an interview with an
2
145, vol. 7, p. 4i.
3
184, vol. 1, pp. 387-388.
informant. About the year 987 (377) he interrogated a

Christian monk who had spent some time in China and he re-
produces this conversation in his Fihrist* Needless to say, this
more direct approach produces not only a vivid presentation of
the freshly gained knowledge, but the information itself is
new. Thus we learn for the first and only time how the "phi-
losophers and wise men of China" 5 imagined that the strange
figures in the horn originated. It is the first impressions of
the newly born animal which are reproduced in the horn. This
explanation of the mythical designs is just another myth, but
we have here at least an awareness of a phenonemon which
needs to be interpreted. 6 Al-Nadim, unlike all the other Mus-
lim writers, thought it worth while to reproduce this informa-
tion. In this instance it is also possible to come fairly close
to the original Chinese version of the According to myth.
one Li Hsiin, it was believed that "the rhinoceros, 'commu-
nicating with the sky' during the time of pregnancy, beholds
the forms of things passing across the sky, and these are re-
produced in the horn of the embryo hence the designation :
'communicating with the sky.' " 7 In both accounts the figures

in the horn are thus due to visual sensations either of the
pregnant mother or of the newly born calf, though the Chinese

author (in his effort to explain the curious name of the ani-
mal) restricts the impressions to heavenly bodies. In view of
this basic agreement it can be assumed that the particular ver-
sion reproduced by al-Nadim was also current in the Far East
and picked up there by his informant. Another unusual state-
ment is contained in al-Nadiim's reference to the price of rhi-
noceros horn. In contrast to the stereotyped wording in all
other Muslim texts mentioning the high cost of the Chinese

4
192, vol. 1, p. 349; translation, 98, vol. 1, pp. 129-130.
5 Ferrand (98, vol. 1, p. 140) speaks erroneously of "les savants de l'lnde,"
the text has IaJUJIpj I
6 It is based on the principle so common in superstitions and in magic that

"like produces like or that an effect resembles its cause" (105, p. 11).
7
165, p. 147; see also p. 137, footnote 1. Li Hsiin lived in the second half
of the eighth century; he made his statement in an account of the drugs of
southern countries.
girdles, the monk reported that the price for the horn had
actually fallen to a fraction of its former value owing to the
change of taste on the part of the ruling Chinese king. In
view of the alert mental attitude of al-Nadim it is not surpris-
ing that he grasped certain facts even better than did his in-
formant. For instance, when he heard the explanation for
the figures in the horn he immediately exclaimed that this must
be the horn of the karkadann. This the monk denied, since
he had heard another name; laboring under the common de-
lusion that different names meant different animals, he was
unable to see the identity of the two.
The profuse variety of nomenclature tripped even the great
al-Biruni,who was led astray by the reports of travelers.
After having reproduced his fine observations of the Indian
rhinoceros (called ganda by him) he continues: "I thought
that the ganda was the karkadann but a man who had visited
Sufala in the country of the Negroes told me that the kark,
which the Negroes call impila, and the horn of which furnishes
the material for the handles of our knives, comes nearer this
description than the karkadann." In spite of this confusion of
terms al-Biruni must have realized a possible connection be-
tween sharav, ganda, karkadann, and kark, otherwise he
would not have grouped together the separate descriptions of
the real and mythical derivatives of the rhinoceros. In doing
this he escaped the misconstructions of many writers like
Marvazi and al-Damirl who amalgamate various re-
tried to
ports so as to achieve an all-embracing, more or less uniform,
but also confusing account.
The early fourteenth-century writer al-Nuwairi gives the
usual potpourri on the karkadann based on al-Jahiz, al-
Mas'udi, and others, but at least he deserves credit for hav-
ing contributed a critical observation about the African rhi-
noceros in contradistinction to the Indian species. Being, as
he was, a government official (katib) and historiographer he
was, of course, not interested in the physical differences of
the two groups but rather in the variety of tales about the
animal. Thus al-Nuwairi is the only medieval author encoun-
tered by me who, after having stated that the Indian rhinoce-
ros is so terrifying that it lets no other animal graze in its

neighborhood, adds that the Ethiopian variety behaves differ-
ently and mingles with other animals. 8 In stating this he makes
use of his advantageous geographical position as a resident
of Egypt, just as about 300 years earlier al-Biruni had profited
from his Indian experiences. As we have briefly noted before,
the observation of reality in the case of an Egyptian writer
has its parallel in the pictorial arts. In the little carpet of a
miniature of the Maqdmdt of 1337 (738) in the Bodleian
Library (pi. 18, lower) we find for the first time in the tradi-
tional motif of the karkadann's pursuit of the elephant a
realistically rendered rhinoceros. In spite of the smallness of
the available area
we have here in defiance of the tradi-
tional unicorn a representation of the bicorned species with
which an Egyptian could be familiar, while a single-horned
animal from India was nothing to him but a literary allusion.
All later writers preferred quantity of curious information
to critical studies. They did not consult those of their con-
temporaries who had actually seen the animal, nor did they
restrict themselves to al-Biruni's trustworthy information. 9
Incredible reports from various sources were no longer ques-
tioned but were gladly accepted for their intriguing details.
Al-Qazwini is typical of these eclectic writers, although he
still manages to write a fairly well-integrated chapter on the
karkadann. Perhaps the most uncritical author on the sub-
ject was al-Damiri, whose hodgepodge account is character-
10
istic of the decline in scientific attitude. In a not too exten-
11
sive report he repeats himself several times because he used
different sources which had the same elements. At other places
he contradicts statements which he has made a few lines be-
8
201, vol. 7, p. 315.
9
Marvazi quotes al-Biruni extensively, but he adds a great deal of fantastic
folklore from classical and Muslim writers.
10 Besides al-Jahiz, al-Zamakhshari, and Abu 'Umar b. 'Abd al-Barr, whose
names are given in his article on the rhinoceros, he leans heavily on al-Gharnati,
'Awfi, and al-Qazwini whom he does not mention as his sources.
11 Thus he mentions the fight with the elephant three timesand the stories of
the "outside feeding" of the foetus nd of the designs on the horn twice each.
fore, such as first declaring that the karkadann has one horn
12
and then attributing three to him. Valuable as his account
might be to a modern scholar as a source of varied informa-
tion, it shows clearly that the writer no longer tried to digest
and integrate his material. But al-Damiri is not by any
means the first to incorporate blatant contradictions into a
medley on the rhinoceros. About 250 years before his time
Marvazi speaks, in the first section of his chapter on the
karkadann, about the strange behavior of the young animal
while it is still in its mother's womb and how, after birth, it
tries to escape the licking of its dam. Only a few sentences
afterward the same writer reproduces the classical myth that
the whole species consists only of males and that no one knows
how they come into existence. Yet Marvazi does not feel the
slightest necessity to ease the shock of such conflicting state-
ments, or to present at least some kind of explanation for
the different theories. Such contradictory data within a single
text let alone those found in the writings of different au-
thors serve to explain how
karkadann could take on so
the
many different forms and could even have two different shapes
in the same manuscript.
Other sources of confusion were the various animals which
were alleged to have single horns and which were sometimes
identified with the rhinoceros. There is no better example of
the state of mind of writers of these centuries than the fact
that they not only described these different imaginary beasts,
but dealt also with their medicinal value and their lawfulness
as food.
It was in this period of uncritical writings that the artists
started to use the karkadann as a figural motif. There was
no well-established iconographic prototype for a unicorn from
13
old Oriental, classical, and Sasanian times. The karkadann
12 Another contradiction is the length of gestation which is once given as
4 years and then as 3 or 7 years. In addition, the rhinoceros is said to be
100 cubits long and even more, and in another place it is alleged to be smaller
than a buffalo.
13 This generalization is not contradicted by very rare exceptions such as
the old oriental model for the animal on the Berlin Mosul jar or the vague
was therefore quite unlike other motifs such as the "sphinx"

or the "harpy," which were popular in this period. The lack
of an iconographic prototype might appear puzzling in view
of the various animals among the Persepolis sculptures which
could have qualified as models for the karkadann. It may be
that the beast which Ibn al-Balkhi described as karkadann
was too fantastic and unlike the usual bovine type described
by the writers to have appealed to the artists. On the other
hand, the frequently occurring reliefs of the bull attacked by
14
a lion did not come up to the traditional picture of the
karkadann as it had crystallized in medieval writings. It is
true that this animal is a bovine and only one horn is shown,
but this iconographic advantage is offset by the fact that the
animal is which was impossible accord-
overcome by the lion,
ing to the texts. Thus the Achaemenian reliefs of the lion and
the bull are but rarely used as pictorial models for a unicorn
15
scene.
The apparent nonexistence of the unicorn motif in Sasa-
nian art is especially important for the development of the
karkadann design in Muslim times. It can be explained by
the fact that the rhinoceros was very little known in Sasanian
Iran and probably completely unknown in earlier periods. The
word karkadann or a similar form does not occur in the extant
Middle Iranian literature. It is not found in the Pahlavi
Bundahishn (in the chapter about the creation of the ani-
mals), nor, for that matter, in the Avesta. There exists also
no other term meaning "rhinoceros" in this literature. 16 In
spite of this lack of direct information the existence of karka-
affinity between the karkadann killed by Iskandar in the Demotle Shdh-ndmah

and the griffin mount of Anahita on a Sasanian silver plate (228, pi. 116).
14
228, pi. 21.
15
For one of the rare exceptions see plate 45 and its explanation on page 69.
Usually the slain animals show two horns or two antlers.
16 Kind information of Prof. B. Geiger. Fie also pointed out to me that
kark occurs in Pahlavi only in the meaning "hen" and in the Avesta only in
the compound kahrkasa "vulture" (literally: hen-eater). Another modern
Persian word for rhinoceros arj which is identical with Pahlavi arz,
occurs in the Bundahishn, though it is counted there among the fish (276,
p. 51).
ii
i5o Freer Gallery of Art Occasional Papers Vol. i
dann in Pahlavi can be inferred with a high degree of certainty

since this word is mentioned as the Persian designation in the
Syriac translation of the Pseudo-Callisthenes. 17 Noldeke has
shown that this translation was most probably made by a
Nestorian at the end of the Sasanian period (seventh cen-
18
tury). This would naturally imply that karkadann existed
in Pahlavi, having been borrowed from the Sanskrit, and that
the Syrian Christians of the Iranian empire became familiar
with this term and used it in their literature. Although this
deduction indicates the use of karkadann in Sasanian times,
the term must have been uncommon, as demonstrated by its
absence in the Pahlavi literature. While the rare use of the
word seems to imply that the people were ordinarily unaware
of the real animal, it should be pointed out that the notion of
a unicorn as such was not unknown in Pahlavi literature. The
Bundahishn speaks of a three-legged, six-eyed, and nine-
mouthed ass whose head is adorned with a single golden,
thousand-branched horn. 19 It stands in the ocean and with its
horn "vanquishes and dissipates all the vile corruption due to
the effort of the noxious creatures." These details indicate that
the concept of this mythical animal is quite different from those
held of the karkadann, which, in spite of fantastic elabora-
tions, always kept a core of reality. The lore of the three-
legged ass had therefore no influence on Muslim literature or
Muslim iconography. 20
In the absence of an established iconographic tradition in
pre-Islamic and early Islamic times, one wonders what it was
that caused the artists to take up the motif and then, at least
17 The
215, Syriac text, p. 211, line 15, translation, p. 119. rhinoceros occurs
there in the corrupt forms marqedad and bargedad. Eighty of the animals were
sent with other gifts by Queen Kundaqa (Candace) of Samraye to Alexander
the Great.
18
199, PP- 13-17-
19
276, pp. 67-69.
20 The "purification of the water through the unicorn" was, however, intro-
duced into the Physiologus. It appears there in the "water-conning" scene,

which later found charming expression in late fifteenth-century tapestries
(242, pp. 60 and 235-236, pi. 5). The Islamic world did not accept this part
of the Physiologus, just as it had rejected the original Iranian myth of the
three-legged ass.
in the East,only as late as the middle of the twelfth century.

Further research might, of course, produce earlier examples,
but they would hardly represent a common type. So far
there is good evidence that no such animal designs were widely
known in the ninth century, since al-Jahiz could refer only to
representations of the 'anqa' and not of the karkadann when
he wanted to prove that the two animals were not identical
21
(see p. 33). That from the late twelfth century on the ani-
mal enjoyed a certain popularity for about 150 to 200 years,
isdemonstrated by the fact that decorative schemes display
it then as an isolated motif, that is, on its own merits and
without any further folkloristic context.

seems that the time has not yet come to give a definite
It
answer for the raison d'etre of the design. The karkadann is
only one of the animals and one of the many motifs popular
at this period, would be necessary to trace the back-
and it
ground of at least the most important ones to find out what

appealed to a Muslim in each of them. However, certain con-
siderations come to mind which might give some explana-
tion why artists were interested in this motif.
Shortly before the first pictures of the karkadann appeared
in the East, the amalgamation of the khutu horn with that
of the karkadann had taken place. This meant that the karka-
dann was no longer regarded merely as the supplier of the
high-priced horn for the curious girdles of the Chinese, which
had been its greatest distinction; its antidotal quality then
made it significant for the Muslims themselves. On the other
hand, the mysterious khutu horn had finally found a body to
21 The only early Islamic example known to me from the literature is a
fragmentary statuette of the animal, excavated in a Khwarizmian castle, said to

be from the end of the eighth century A.D. Representations of a four-armed
deity from a contemporary castle of the same region indicate cultural rela-
tions between Khwarizm and India at this period which could also explain
the occurrence of the rhinoceros in this early art (101, p. 164). The only
statuette of a rhinoceros mentioned Muslim literature with which I am
in the
familiar is the one of gold and jewels that was given by native chieftains to
Iskandar after he had spoken to the "talking tree." This precious object was
given with other gifts, among them two elephant tusks, thus possibly indicating
an Indian locale (102, vol. 4, pp. 232-233, line 1575; 103, vol. 6. p. 169).
be attached to and could thus be represented in an animal

decoration. This must have made the Muslim world receptive
to the inclusion of the karkadann as a motif at this particular
time.
Furthermore, an additional reason for its new popularity
was possibly the sequence of animals chasing each other, which
constituted a favorite Seljuk decoration for circular bands on
metal and pottery objects. As the fiercest beast and killer of
all animals, including the elephant, it was an obvious choice
for this decorative scheme. Finally, it should be remembered
that this period was very fond of fantastic animals such as
those usually defined as sphinxes, harpies, and griffins. The
karkadann fitted well into this group.
The era was receptive to becoming visually aware of new
a
animal. The commercial activities of the big cities made a
large body of learned men These scholars and their
possible.
colleagues, supported by princes, were exhaustive in their en-
cyclopedic surveys of the world, its people, animals and plants.
This is testified by the works of al-Idrisi, Yaqut, Ibn al-Baitar,
al-Qazwini, Ibn Khallikan, and others, all of which reveal a
tremendous curiosity toward the various aspects of the

world even of its "wonders"
22
and an attempt to satisfy
this widespread yearning. What these learned books might
lack in originality they made up by a great deal of detailed
information, real or fancied. There is not only an insistence
on the correct dates and spelling of names, but also on visual
recording. Thus when Ibn al-Suri (died 1242/639) was
botanizing in Syria he was accompanied by an artist who made
colored drawings of plants in the different stages of their
23
growth. It was also at this time and in line with this thor-
oughness that the illustration of manuscripts dealing with
animals and plants was deemed necessary. This learned ac-
tivity extended even to imaginary monsters and supernatural
beings from a lower folkloristic level. One has only to look
22 and 2 passim.
234, vol. 2, pts. i
23
234, vol. 2, pt. 1, p. 54; vol. 2, pt. 2, p. 649, quoting Ibn Abi Usaibi'a (130,
vol. 2, p. 219).
through illuminated copies of al-Qazwini's text to see the

many strange creatures which existed in the world of a
thirteenth-century Muslim.
All these factors seem to explain the existence of the karka-
dann in Muslim art. If there is any doubt left, it might center
around the isolated figures of the animal. Here the folk-
loristic appeal of the elephant-hunt scenes with their delight
in telling a story does not apply; and we can likewise assume
that the artist did not haphazardly select this unusual animal
just to satisfy scientific curiosity. Was it then the magic
quality of the horn that caused the artist to prefer a fearful
monster to other equally decorative designs with more pleasant
associations and led him to apply it several times to a beauti-
ful luxury object (pis. 1 and 2) ? Or had the design become
common enough so that not too much attention was paid to
it as an individual motif? If we knew more about the work-
ing of the Muslim mind in the twelfth, thirteenth, and four-
teenth centuries, the answer to these questions would not be
so difficult. One factor which prevents a solution is our present
lack of knowledge as to whether there is any psychological
basis for the choice of certain motifs in Muslim art. We thus
do not know whether we are confronted with a case parallel
to certain phenomena in Western art. During the Romanesque
period, for instance, one frequently finds capitals and initial
letters decorated with monsters which do not fit into the usual
religious aspect of contemporary art. These creatures have
been explained as "psychologically significant images of force,
aggressiveness, anxiety and fear;" in other words, as "pro-
24
jected emotions." It seems still premature to speculate
whether the fascination of the Muslim artist for a violent all-
powerful monster, which without mercy kills man and beast
and is only rarely vanquished and subdued, is likewise condi-
tioned by human fears. As has been stated before, the final
explanation of the motif lies in the further study of all impor-
tant designs used during the Middle Ages.
Once the picture of the karkadann was presented to the
236, pp. 132-137-

public its appeal lay in evoked a wide response.

the fact that it
It proved equally attractive to the scholar and to an illiterate

audience of popular stories; only their intellectual reactions
and mental associations were different. It is also certain that
whatever the intent of the artist and the education of the on-
looker may have been, neither of them seems to have been
concerned with a conscious symbolism in the design.
In view of the lack of an established iconographic proto-
type, medieval Muslim artists had to rely for their representa-
tions mainly on the writers. Naturally they were influenced
by the opinion of their contemporaries like al-Qazwini and
al-Damiri though they also followed long-established descrip-
tions such as those of al-Jawhari and Firdawsi. All these
authors proved to be difficult sources of information. The
texts were often fantastic and contradictory and their termi-
nology was misleading. The descriptions provided material
for only some of the bodily aspects of the animal, while a
great deal of intriguing information was of no use to a painter
or sculptor. One cannot blame the writer for restricting his
accounts because his aim was not to provide the raw material
for a decorative artist; yet these limited and fanciful sketches
were all that the painter had to go on for his designs.
In view of this handicap the artist needed further help. In
some instances he was able to exploit a peculiar linguistic con-
dition, because by reading J? to mean gurg, "wolf," all
iconographic difficulties were removed and he could then easily
paint a wolflike animal with a single horn. I am not familiar
with other Muslim examples of substitutional iconography
based on two meanings of a written word, but the phenomenon
25
has, for instance, been observed in India.
-3
See the Indian motif of the elephant-carrying bird which was originally
the solar bird garuda carrying in its talons or beak the chthonic snake naga.
Since naga means both snake and elephant the one animal was substituted for
the other and a new iconographic type was created (127, pp. 17 and 21; 265,
vol. 2, letterpress to pi. 59; 279, pp. 255-257). In this case the substitution is
due two definitions of a word, while in the case of karg-gurg, we have

to the
not only two definitions but also two pronunciations. By the way, the elephant-
carrying garuda is also the prototype for the elephant-carrying simurgh, see
p. 32 and pi. 17.
No. 3 I. The Unicorn Ettinghausen T55
Muslim art provides, however, another instance where the

re-interpretation of the name created the physical appearance
of an animal unknown to the artist. This happened in the case
of the giraffe, called in Persian shutur-gav-palang c^JJL> jtf J
(literally, "camel-ox-leopard"), illustrated in the al-Qazwini
manuscript of 1789 in the Freer Gallery of Art. When the
Indian artist had to reproduce this unfamiliar animal he did
not follow the text (which starts with a statement that the
head is like that of a camel, and so on), but he concocted a
creature whose bovine-horned head is carried on a long neck
above the furry body of a leopard (pi. 16, upper). 26 Here

too, then, there is hardly any doubt that it was the word
which created the physical form.
Substitutional iconography engendered by linguistic condi-
tions was by its very nature only of limited help. In the case
of the unicorns was applied only to some of the animals in
it
the Shdh-ndmah illustrations. For other texts and other occa-

sions the artist needed inspiration for his imagery. In
still
certain instances he eagerlyemployed foreign iconographic

models. This happened when these models seemed to fit the
mental picture he had formed from the scant information in
Muslim literature. To this category belongs the buffalo- or
cowlike karkadann which follows an iconographic type from
China; or we can refer to the karkadann in the shape of an
antelope, deer, or goat which ultimately seems to have been
derived from Indian models, literary or figurative. It is also
for this reason that an iconographic model was assumed to
be the most likely prototype of the equine karkadann and not
alone the occasional hints about a relationship of the karka-
dann with the horse, as found in Arab authors.
The happy conjunction of textual allusions with foreign
iconographic types which gave visual realization to several
26 There is nothing of a camel in this "giraffe," if we do not regard the

elongated neck as having been inspired by that animal.
The process which leads from a descriptive name to a new type of animal
is also found, in China, in the case of the Chinese word t'o ni'ao or ostrich,
actually "camel-bird," which is represented as the literal meaning of the
word implies (165, pp. 126 f., figs. 16 and 17).
forms of the karkadann is, of course, not an isolated phe-

nomenon. It was similarly fecund in the case of other animals.
Thus the simurgh turned, in the Mongol period, from a mere
parrotlike bird into the much more colorful and more awe-
inspiring feng-huang type of Chinese derivation, which from
that time on graced so many fine miniatures and other art ob-
jects.
27
Another good example is the faras al-ma' *Ul oV> 28
whose name means literally "water-horse," but obviously re-
fers to the hippopotamus. This animal is described as an
aquatic "superhorse" in such zoological treatises as the 'Ajd'ib
al-makhluqdt or the Manafi'-i hayavan. 29 It required the ap-
pearance of the Far Eastern horse with flames emanating
30
from above the front legs (pi. 48, upper) to create in the
Mongol period a picture of the animal which was then ac-
cepted as a proper rendition of the animal (pi. 48, lower). 31

The Muslim artist likewise used foreign designs to illus-
trate certain details in the animal physique. Thus when the

chinoiserie trend of the sixteenth century introduced Chinese
27 The early type is represented in a Mu'nis al-ahrdr miniature (pi. 8), while
the Far Eastern type is found, for instance, in the Sulaiman-Bilqls and other
paintings (pis. 34-37).
28 Another Arabic name is faras al-bahr jsn^ <j" j3 "horse of the sea." In
\
Persian the animal is called asb-i abi ^> j _~|.
20 "One says it resembles the horse on land, only that its mane and tail are
bigger, etc." (al-Qazwini.) The same author reports also a story of a dark-
colored horse with white dots which came out of the water and covered a
mare (217, vol. 1, p. 141; 218, pp. 228-229). The Mandfi'-i hayavdn in the
Morgan Library states "its face and forehead resemble those of a horse"
(5, fol. 29a).
30 See also and 129 (mirrors of the Sui and
261, pt. 2, vol. 2, pis. 117
T'ang periods). No examples of the design from the Sung and Yuan periods
are known to me, but Dr. Cammann informs me that the antiquarians of the
Sung period knew the motif on the T'ang mirrors and called it hai-ma "sea-
horse." This legendary animal with its flames above the front legs was used
in mandarin squares to denote a low military rank (68, p. 110, fig. 11b). Ac-
cording to Cammann it is first mentioned for this use in the specifications for
Ming military insignia of 1393.
31 See also 157, fig. 14. The Sarre manuscript of al-Qazwini shows, however,
a winged horse (154, vol. 5, pi. 853B). This is a parallel to the representations
of the winged karkadann. In both cases the wings seem to imbue the animals
with the special qualities mentioned in the text.
fabulous animals with branched and crooked horns (see pis.

43, upper, and 45) this type of horn with its "protuberances"
was applied to the nose of the karkadann in a Beatty manu-
script (pi. 14, upper) since it seemed to tally with the de-
scription in the al-Qazwini text.
In taking over a foreign iconographic model it was to be
expected that the coloration would be changed or certain ana-
tomical details would be misinterpreted or altogether dis-
carded. We noticed this in the case of the Princeton
shadhahvar, which
32
is red not white, as a hsieh-chai type ani-
mal should be. It is likewise quite natural that the Near
Eastern artist was entirely unaware of the symbolical mean-
ing of the animal that he copied. It thus happens that in Islam
the animal carries with mental associations different from
it
those of its prototype. In China the ch'i-lin is the noblest and

most perfect of all animals, the emblem of goodness and
33
virtue, and its apearance a sign of happy augury, while the
karkadann, which at least in one case took over its shape, is
a ferocious and tyrannical beast. The same contrast is to be
found in the reinterpretation of the hsieh-chai as shadhahvar.
The Chinese monster stands for equity, since it is able to dis-
tinguish between right and wrong, 34 while the Near Eastern
animal is characterized by its sound-producing horn, although
some writers have also referred to its carnivorous appetite.
In spite of all the help given by literary references and
foreign models, artistic imagination was still necessary and
the artists provided it in many of their creations. This re-
sulted in many "sports" which were, like all mythical crea-
tions of the East, "zoologically beguiling" and imbued with
lifelike appearance, though nothing similar to them existed in
nature.
The great disparity of iconographic types until recent times
naturally makes one wonder why the Islamic artists did not
continue to paint the rhinoceros realistically, once they had
determined what the animal was like and were able to por-
32 68, p. 108.
33
273, t. 1, p. 423.
34
68, p. 108.
tray it Applied to specific representations

in its actual shape.
this question could also be phrased: how can it be explained
that the early seventeenth-century artist illustrating the SJidh-
ndmah Metropolitan Museum showed the karg as an
in the
antelope with a horn on its nose, after earlier artists, like the
painter of the Princeton Shdh-ndmah, had been able to pro-
duce the general impression of the animal and all its charac-
teristic features?
The answer seems to be twofold: the condition is due both
to a mental attitude and the lack of a technical process. There
is first the continued uncritical belief in early authorities whose
writings were handed down through generations, often aug-
mented or reduced in later compilations but not basically
changed, let alone improved, by new observations. This ex-
plains the continued use of the al-Qazwini text with its many
absurd myths of the karkadann, sinad, harish, and shadhahvar,
even after many people had become aware that the stories
could not be true, and it accounts likewise for the adherence
to traditional iconographic types, after the true character of
a natural phenomenon had been established. This intellectual
authoritarianism has often been quoted as a cause of cultural
stagnation. Less obvious is the second factor, namely, the
technical inability to make number of unvarying copies
a large
of pictures same text. This prevented the
all illustrating the
wider dissemination of new knowledge based on more dis-
criminating observation, and thus made scientific work in the
modern sense impossible. With this limitation Muslim
scholars suffered the same handicaps in scientific reporting as
the writers of antiquity and the Middle Ages. The West
overcame this disadvantage in about 1461 when the first illus-
trated printed book, Boner's Edelstein, was published in Bam-
berg, and was soon to be followed by other books with figures
35
of tools and well-observed natural objects. Such scientific
reporting, with its wide distribution of immutable pictures,
could not be paralleled in Turkey before 1729 when, as the
second printed Turkish book, a history of the Ottoman navy,
35
140, pp. 54-56.
with maps, was printed in Istanbul; 36 nor in Egypt until

five
1800 when, during the French occupation, the first all-Arabic

37
publication, the journal Tanbih, was printed in Cairo; nor
in Iran until 18 16, when the first press was introduced in
Tabriz. 38 The and 1729, 1800 and 18 16, illustrate

dates 146 1
in a nutshell the difference in scientific development in the two

civilizations.
Since there never seems to have been a great demand for
representations of the karkadann, no definite iconographic
type crystallized to become universally accepted throughout
the Islamic world. Only the bovine type came close to this
distinction. In Spain a decided preference for one type was
evolved, otherwise no regional preference for one or the other
version is recognizable. The Spanish version, however, pre-
sents a special case, since the winged feline karkadann with
a short stubby head and a long curved horn with helical
grooves is, as far as we know, to be found in but one medium,
ivory,and only during the first half of the eleventh century

A.D. In the Islamic world outside al-Andalus, the literary
and iconographic sources used by the artist and his individual
approach to the artistic problem is usually more important in
understanding a type than the country of origin of the decora-
tor or the region where he worked.
While the karkadann occurs fairly frequently in Syria and
Iran, and is likewise to be found in Spain, Iraq, and Anatolia,
one wonders why Egypt and India, the two countries which
were in a position to be better informed about the animal than
others in the caliphate, did not make wider and more original
use of it. The rarity if not lack of the design in early Egyp-
tian decoration can be explained by the fact that the animal
36 The book by Hajji Khalifa entitled Ttihfat iil-kibdr ji esfdr
31, p. 12. is
ul-bihdr ("Present for the Great Ones Dealing with Naval Wars"). In 1729-
1730 (1142) a book of fables appeared entitled Ta'rikh ul-Hind iil-gharbi ("His-
tory of the West Indies"), which contained 4 geographical maps, 1 celestial
chart, and 13 figures of men, animals, and plants (31, p. 14).
37 After the departure of the French in 1801,
107, p. 149 (No. 16 of his list).
printing in Egypt stopped until 1822 when, on orders of Muhammad 'All, it
was again introduced into the country (107, p. 157; 42, pp. 13 If.).
38 61, vol.
4, pp. 155 and 468.
i6o Freer Gallery of Art Occasional Papers Vol. i
came vogue only after the middle of the twelfth

into general
century. It is thus apparently lacking in the Fatimid reper-
39
tory. In the Mameluke period it occurs mostly on brass ob-
jects, which, though made in Cairo, reveal a great deal of
Mesopotamian inspiration, if they were not in certain cases

actually made by Mesopotamian artists. 40 Since the motif on
Egyptian pieces of the karkadann attacking other animals is
no different from earlier representations of the same subject
on objects made in regions farther east, it may be surmised that
the design came to Egypt from this part of the Muslim world.
Only where the artists were on their own (as in the case of
the carpet designs of plate 18, lower) did an original version
result, which took notice of the specific character of the African
rhinoceros.
Equally surprising, on first thought, is the not too frequent
use of the motif in India, where the animal is at home and
must have been known to many artists. It is not represented,
for instance, on the ivory powder primers of the Mughal
41
period which show combinations of all sorts of animals. It
is likewise missing on a large Mughal carpet in the Textile
Museum in Washington on which many different specimens
of Indian fauna are realistically rendered in a landscape
42
setting. There is only a vague stylization of it on the animal
rug with fantastic zoological combinations. 43 However, the
carpet in the National Gallery in Washington and several
39 176. Rhinoceros horn is not mentioned in al-Maqrizi's account of the
treasures of the Fatimids (150). The horn already had, however, a reputation
in Pharaonic times, owing to its healing and magical powers. In the tomb of
Hor-Aha pottery imitations of the horn were found which served as substitutes
for real pieces and were supposed to have the same magical powers. The
animal had disappeared from Egypt in historical times (249, pp. 42 f., foot-
note 4).
40 Thus a tray in the Metropolitan Museum of Art (No. 91.1.602) was made
in Cairo by one Husain b. Ahmad b. Husain of Mosul (84, p. 234 and fig. 3).
It is closely related to another tray in the same museum (No. 91. 1.605) which
has three representations of the motif of the attacking karkadann (see above,
p. 28, footnote 9).
41
51, figs. 1-9, 11-12.
42 No. R 6.3. This carpet, which is nearly 20 feet long, is still unpublished.
43 and
15, title picture fig. 3.
Mughal miniatures prove that the rhinoceros in its actual

form is not altogether foreign to the artistic repertory of
Muslim India. Its apparent rarity conforms only to the scant
interest paid to it in the non-Muslim literature and arts of
India.
The decline of the decorative arts in the Muslim world
since the time of Timur, noticeable in pottery, glass, and even
metalwork, had its effect on the use of the karkadann motif.
The imagination of the artists in the twelfth, thirteenth, and
early fourteenth centuries was no longer as vivid as before.
The craftsmen seem to be satisfied with certain standard
shapes and themes of ornamentation. This eliminated many
decorative motifs, among them the isolated karkadann and the
karkadann-elephant fight. The fact that al-Damiri no longer
mentions the antidotal quality of the karkadann horn, in spite
of the fact that he was familiar with al-Qazwinl's text, also
helps to explain their disappearance. The animal continues,
however, to be painted in the always popular 'Ajaib al-
makhluqdt and Shdh-ndmah manuscripts.
The tradition of the legendary karkadann, based on book-
ish and uncritical literary sources, was broken only in the
44
early sixteenth century. This must have been due in the first
44 This fact has some bearing on the critical evaluation of objects. It applies
for instance in the case of a large green-glazed Persian bowl, formerly in the
V. Everit Macy collection (82, pp. 8-9, No. 30; 172, p. 124, No. 488, illustrated).
The vessel belongs to a ceramic group which has been attributed to the Garrus
district in Kurdistan (213, vol. 2, p. 1531), and it has been dated tenth to
eleventh century. It is decorated with the figure of a rhinoceros easily dis-

tinguishable by many of its characteristic features such as the bulky mass of
its body, the pointed ears, the wide mouth with the protruding upper lip, and
especially the curved horn on its nose. How can such a representation be
explained, when in spite of diligent search it has not been possible to find
another representation of a rhinoceros dating from before 1337 which approxi-
mately reproduces its physical appearance or shows at least the correct posi-
tion of the horn? It will be readily admitted that this writer knows only a
limited portion of theMuslim objects which portray the "unicorn"
existing
and that these represent only a minute fraction of all the designs made in
the Middle Ages. Still, it must also be taken into account that popular, scholarly
treatises such as those of Ibn Bukhtishu', al-Qazwini, and al-Damiri knew
nothing of the most obvious feature of the animal, the position of its horn.
Furthermore, the motif was hardly known when the Garrus pottery was
made. Yet, a village potter working in a provincial district of Iran, a great
place to the fact thatmore people became familiar with the

animal so that information about its actual appearance could
spread in the Muslim world. In this respect it will be remem-
bered that from the late fourteenth century on there exist
records of rhinoceros hunts by princes and rulers. The possi-
bility of direct observation was artistically exploited when, in
the late fifteenth century or even more so in the sixteenth cen-

tury, the trend in Persian miniature painting turned toward
realism. This explains the sudden appearance of a fairly well-
observed rhinoceros in the otherwise undistinguished Shah-
nam ah of 1544 (951), while the manuscripts of the second
half of the fifteenth century had still used the legendary type
of the karkadann (pis. 26, 27, and 30).
This trend in Persian miniatures coincides with the rise of
the Mughal empire as a leading Muslim power, which meant
also the rise of a new school of painting preponderantly real-
istic in its nature. Most of the realistic renditions of the rhi-
noceros thus come from India, which is not only a home of the
animal but, in the Mughal period, is also the source of remark-
ably lifelike portraits of animals. On the other hand, Persian
artists often produced strange hybrids in which a new percep-
tion was curiously blended with traditional ideas.
In the eighteenth century, al-Qazwini manuscripts of in-
ferior quality, and thus destined for the simple and impecuni-
ous, showed illustrations of the karkadann, in which a kind of
dreary resemblance to the rhinoceros emerged. 45 The text, of
course, still tells the old tales and superstitions, but the minia-
tures have now nearly caught up with the actual animal. The
encounter with reality is, however, disenchanting. The fero-
cious and yet impressive character of the old monster has gone
and all that remains is an immense and unprepossessing hulk
of a body. No new ramifications of the age-old myth could
possibly grow up around this sort of an animal.
distance away from the regions where the rhinoceros lives, is said to have
made the most realistic representation of the animal in the early Muslim Middle
Ages. The only conclusion to be drawn from all these facts seems to be that
either the decoration of the bowl was changed when the piece was restored
or that the bowl is not as old as was hitherto believed.
45 See 9, fol. 112a; and 10, fol. 465a, old collation (pi. 14, lower).
LITERATURE CITED
References in the text are made by citing the bold-face numerals before
each title.
Manuscripts:
[Author Unidentified.]
1.
'
Ajd'ib al-makhluqat wa-gharaib al-mawjuddt
ob^^-Jl j&3 O^J^Jl (in Persian). Written in the
late sixteenth or seventeenth century A.D. following an older,

fifteenth-century prototype. Walters Art Gallery, Baltimore,
ms. W. 593.
Ibn Bukhtishu', Abu Sa'id 'Ubaid Allah b. JibrIl, and Aristotle.
2. Nat al-hayawdn wa-mandfiuhu O^^-l C (in Ara-

bic). Written in the thirteenth century A.D. British Mu-
seum, ms. OR. 2784.
3. Manaft al-hayawdn O^^-l (in Arabic). Written in 700
(A.D. 1300). Bibliotheque Nationale, Paris, ms. Ar. 2782.
4. Taba'i al-hayawdn wa-khawdssuhu <u?l^\$ 0^^*1 ^M* ( m

Arabic). Written probably in the fourteenth century A.D.
Princeton University Library, Robert Garrett collection, ms.
No. 203 B.
Ibn Bukhtishu', Abu Sa'id 'Ubaid Allah b. Jibril (?).
5. Mandfi-i hayavdn 9 ^*Persian). Written in Ma-

(in
f
ragha in 694 or 697 or 699 (A.D. 1294 or 1297 or 1299).
The Pierpont Morgan Library, New York, ms. M. 500.
Marvazi, Sharaf al-Zaman Tahir.
6. Taba'i al-hayawdn b\^-\ jfM* ( m Arabic). Probably writ-
ten in Iran by an Iranian in the first half of the twelfth cen-
tury. India Office Library, ms. Delhi, Arabic 1949.
7. Taba'i al-hayawdn
fM* m ( Arabic). Thirteenth cen-
tury? British Museum, ms. Add. 21.102.
Al- Qazwini, Zakariya' b. Muhammad.
'
8. Ajd'ib al-makhluqat wa-gharaib al-mawjuddt i*J\s>zC>
ob^>Jl ^J\j&3 O^y^Jj (in Arabic). Written in the
second half of the fourteenth century. (The so-called "Sarre

manuscript.") Private collection.
163
g. 'Ajaib al-makhluqdt wa-gharaib al-mawjuddt 5

'
v**^*
ob^\$-Jl ^J\j^3 O^y^tJl (in Turkish). Written by
Muhammad b. Muhammad in 1121 (A.D. 1709). Walters
Art Gallery, Baltimore, ms. W. 659.
10. 'Ajd'ib al-makhluqdt wa-ghard'ib al-mawjuddt V^*^
ob^-^Jl *J>\j&3 o^y^Jl (in Persian). Finished 5th
Dhu'l-Qa'da of the year 31 of the reign of Shah 'Alain (A.D.
1789). Freer Gallery of Art, No. 07.625.
Books and Articles:
A bhinishkramanasutra.
11. The romantic legend of Sakya Buddha from the Chinese-San-
scrit. By Samuel Beal. London, Triibner & Co., 1875.
Ackerman, Phyllis.
12. The talking tree. In American Institute for Persian Art and
Archaeology, Bulletin, vol. 4, No. 2, pp. 67-72. New York,
December 1935.
13. Three French Renaissance tapestries. In Gazette des Beaux-
arts, 6th ser., vol. 29, pp. 207-212. New York, 1946.
Aelianus, Claudius.
14. Claudii Aeliani De natura animalium libri XVII, Varia
historia, Epistolae fragmenta, et recognitione Rudolphi
Hercheri ... 2 vols. Lipsiae, in aedibus B. G. Teubneri,
1864-66. (Bibliotheca Scriptorum Graecorum et Romanorum
Teubneriana [S.g.].)
Aga-Oglu, Mehmet.
15. A fragment of a rare Indian carpet. In Detroit Institute of Arts
of the City of Detroit, Bulletin, vol. 13, pp. 1-5. October,
I93I-
16. Persian bookbindings of the fifteenth century. Ann Arbor, Uni-
versity of Michigan Press, 1935. (University of Michigan
Publications. Fine Arts, vol. 1.)
Akhbdr al-Sin wal-Hind.

17. (j^l 'Ahbdr as-Sln wa-l-Hind. Relation de la
Chine et de VInde, redigee en 851. Texte etabli, traduit et
commente par Jean Sauvaget. Paris, Societe d'Edition "Les

Belles Lettres," 1948.
18. Relation des voyages faits par les arabes et les per sans dans
VInde et a la Chine le IX e
siecle de Vere chretienne. Texte
arabe imprime en 181 1 par les soins de feu Langles publie avec
des corrections et additions et accompagne d'une traduction
franchise et d'eclaircissements par M. Reinaud ... 2 vols.
Paris, Impr. Royale, 1845.

19. Voyage du marchand arabe Sulayman en Inde et en Chine
redige en 851. Suivi de remarques par Abu Zayd Hasan, tr.
par Gabriel Ferrand. Paris, Editions Bossard, 1922. (Les
Classiques de l'Orient, vol. 7.)
A If laila wa-laila.
20. iLJj uaJ I (The Arabian nights) . 4 vols, in 2. Cairo, 1354

H./ A.D. 1935.
21. Die Erzdhlungen aus den tausend und ein Nachten ... 6 vols.
Calcuttaer Ausgabe vom Jahre 1839 iibertragen von Enno
Littmann. Leipzig, Insel-verlag, 1923-28.
22. The thousand and one nights; commonly called in England, The
Arabian nights' entertainments. A new translation from the
Arabic, with copious notes, by Edward William Lane. Illus-
trated by many hundred engravings on wood, from original
designs by William Harvey. 3 vols. London, Charles Knight
and Co., 1839-41.
23. The thousand and one nights; The Arabian nights' entertain-
ments. Tr. by E. W. Lane, ed. by Stanley Lane-Poole. 4
vols. London, George Bell and Sons, 1906.
'AllamT, Abu 'l-Fadl.
24. The Ain i Akbari, by Abul Fazl 'Allami, tr. from the original
Persian by H. Blochmann [and H. S. Jarrett]. 3 vols. Cal-
cutta, printed by G. H. Rouse at the Baptist Mission Press,
1873-1894.
Andersson, Karl Johan.
25. hake Ngami; or, Exploration and discoveries during four
years' wanderings in the wilds of southwestern Africa. New
York, Dix, Edwards & Co. ;
London, Hurst & Blackett, 1857.
Anecdota Syriaca.
26. Anecdota Syriaca; collegit, edidit, explicuit Jan Pieter Nicolaas
Land. 4 vols. Lugduni Batavorum, E. Brill, 1862-1875.
Aristotle.
27. Historia animalium, tr. by D'Arcy Wcntworth Thompson.
Oxford, Clarendon Press, 1910. (The works of Aristotle, tr.
under the editorship of J. A. Smith and W. D. Ross, vol. 4.)
12
28. Parts of animals, with an English translation by A. I. Peck and

a foreword by F. H. A. Marshall . . . Cambridge, Mass.,
Harvard University Press ;
London, W. Heinemann, Ltd.,
1937. (Loeb Classical Library. Greek Authors.)
Arnold, Sir Thomas Walker.
29. Painting in Islam . . . Oxford, The Clarendon Press, 1928.
Asadi, Tusi, the Younger.

30. Asadi s neupersisches Worterbuch, Lughat-i Furs, nach der
einzigen vaticanischen Handschrift herausgegeben von Paul
Horn. Berlin, Weidmannsche Buchhandlung, 1897. (Ab-
handlungen der Koniglichen Gesellschaft der Wissenschaften
zu Gottingen. Phil.-hist. klasse, n. f., Bd. I, Nr. 8.)
'Awfi. See Nizamu 'd-din.
Babinger, Franz.
31. Stambuler Buchwesen im 18. Jahrhundert. Leipzig, Deutscher
Verein fur Buchwesen und Schrifttum, 19 19.
Babur, Emperor of Hindustan.
32. The Bdbur-ndma in English {Memoirs of Babur), tr. from the
original Turki text of Zahiru'd-din Muhammad Babur
Padshah Ghazi by Annette Susannah Beveridge ... 2 vols.
London, Luzac & Co., 1922.
Bacher, Wilhelm.
33. Nizami's Leben und Werke und der zweite Theil des Nizd-
mischen Alexanderbuches . . . Gottingen, Dieterichsche Uni-
versitatsbuchdr., 187 1.
Baer, K. E. von.
34. Anatomische und zoologische Untersuchungen iiber das W all-
ross (Trichecus rosmarus) und Vergleichung dieses Thiers
mit andern See-Sdugethieren. [St. Petersburg? 1838.] (Aus

den Memoiren der Kaiserlichen Akademie der Wissenschaften
besonders abgedruckt.)
Barthold, W.
35. Bulghdr. In The Encyclopaedia of Islam, vol. I, pp. 786-791.
Leyden, E. J. Brill, Ltd.; London, Luzac & Co., 1913-38.
Bashkiroff, A. S.
36. The art of Daghestan, carved stones. Moscow, Ranion, 193 1.

(Scientific Research Institute of Ethnical and National Cul-
ture of the People of the East in U. S. S. R.)
Beard (or Bird?), Hubert.

37. j> O Li-mddhd kdna li-wahldi 'l-qarni jildun mu-
jd'adun. 0^1 ^ In al-Mustami
al-arabl. The Arab Listener, vol. 8, No. 12, p. 20, 1947.
Berg, Bengt.
38. Meine Jagd nach dem Einhorn. Frankfurt am Main, Riitten &
Loening Verlag, 1933.
Berlin. Staatliche Museen, Islamische Abteilung.
39. . . . Indische Miniaturen ... 2. aufl. [mit einer Einftihrung
von Ernst Kiihnel]. Berlin, 1937. (Bilderhefte der Isla-
mischen Abteilung, Hft. 1.)
Bestiary, Latin.
40. The bestiary; being a reproduction in full of the manuscript
Ii.4.26 in the University library, Cambridge, with supple-
mentary plates from other manuscripts of English origin, and
a preliminary study of the Latin bestiary as current in Eng-
land; edited for the Roxburghe Club by M. R. James. Ox-
ford, printed for the Roxburghe Club, by J. J. Johnson at the
University Press, 1928.

Bestiary, Latin, 1921.
41. A Peterborough psalter and bestiary of the fourteenth century,
described byMontague Rhodes James Oxford, printed . . .
for presentation to the members of the Roxburghe Club [at

the Oxford University Press, by F. Hall] 1921.
Bianchi, T. X.
42. Catalogue general et detaille des livres arabes, persons, et turcs
imprimis a. Boulac en Egypte depuis V introduction de Vimpri-
mcrie dans ce pays, en 1822, jusquen 1842. Paris, Impr.
Royale, 1843. (Reprinted from: Journal A siatique, No. 14,
1843.)
Binyon, Sir Laurence, Wilkinson, J. V. S., and Gray, Basil.
43. Persian miniature painting, including a critical and descriptive
catalogue of the miniatures exhibited at Burlington House,
January-March, igji . . . London, Oxford University
Press, H. Milford, 1933-
al-Biruni.
44. Alberun'is India; an account of the religion, philosophy, litera-
ture, geography, chronology, astronomy , customs, laivs and
astrology of India about A.D. 1030. An English ed., with
notes and indices. By Dr. Edward C. Sachau . . . London,
K. Paul, Trench, Triibner & Co., Ltd., 19 10.
45. Fragments arabes et persans relatifs a Vlnde, recueillis par M.

Reinaud. Nos. III-IV. Extrait de Vouvrage d' Albyrouny
sur Vlnde. In Journal Asiatique, 4 e ser., t. 4, pp. 22 1 -300.
Paris, 1844.
45a. Kitdb al-jamdhir fl ma'rifat al-jawdhir ^ kJc!>
j*}^ Hyderabad, 1355 H.

Bishop, Carl Whiting.
46. Rhinoceros and wild ox in ancient China. In The China Jour-
nal, vol. 18, pp. 322-330. Shanghai, 1933.
47. The ritual bullfight ... In Smithsonian Institution, Annual
Report of the Board of Regents, pp. 447-455. Washington,
D. C, 1926.
Blanford, W. T.
48. The fauna of British India, including Ceylon and Burma:
Mammalia . . . London, Taylor and Francis, 1888.
Blochet, Edgar.
49. Les enluminures des manuscrits orientaux turcs, arabes, per-
sans de la Bibliotheque nationale . . . Paris, Editions de la
Gazette des Beaux-arts, 1926.
Bond, Sir Edward A., ed.
50. Russia at the close of the sixteenth century . . . London,

printed for the Hakluyt Society, 1856. (Works issued for
the Hakluyt Society, No. 20.)
Born, Wolfgang.
51. Ivory powder flasks from the Mughal period. In Ars Islamica,
vol. 9, pp. 93-1 1 1. Ann Arbor, Mich., 1942.
Boston. Museum of Fine Arts.
52. Les miniatures orientales de la collectionGoloubew au Museum
of Fine Arts de Boston, par Ananda K. Coomaraswamy . . .
avec un avant-propos de Victor Goloubew Paris et . . .
Bruxelles, Les Editions G. van Oest, 1929. (Ars Asiatica;

Etudes et Documents Pub. par V. Goloubew . . . 13.)
Brehm, Alfred Edmund.
53. Brehms Tierleben : Allgemeine Kunde des Tierreichs ... 4.
aufl. von Otto zur Strassen ... 12 vols. Leipzig und

Wien, Bibliographisches Institut, 1911-19.
Brian, Doris.
54. A reconstruction of the miniature cycle in the Demotte Shah
Namah. In Ars Islamica, vol. 6, pp. 97-112. Ann Arbor,
Mich., 1939.
Briggs, George W.
55. The Indian rhinoceros as a sacred animal. In American
Oriental Society, Journal, vol. 51, pp. 276-282. New Haven,
Conn., 1 93 1.
British Museum. Dept. of oriental printed books and manuscripts.

56. Supplement to the Catalogue of the Arabic manuscripts in the
British museum ... by Charles Rieu . . . London, Long-

mans & Co., 1894.
Brockelmann, Carl.
57. Geschichte der arabischen Litteratur ... 2 vols. Weimar
[etc.], E. Felber, 1898-1902.
Supplementband ... 3 vols. Leiden, E. J. Brill,
1937-39-
58. Volkskundliches aus Altturkistan. In Asia Major, Bd. 2, pp.
1 10-124. Leipzig, 1925.
Broom, Robert.
59. The first white rhino in captivity. In Natural History, vol. 55,
pp. 432-433. New York, 1946.
Brown, William Norman.
60. Buddha's teachings, being the Sutta-nipata or discourse-collec-
tion. Edited in the original Pali text ... by Lord Chalmers
[a review]. In American Oriental Society, Journal, vol. 54,
p. 218-219. New Haven, Conn., 1934.
Browne, Edward Granville.

61. A literary history of Persia ... 4 vols. Cambridge, The Uni-
versity Press, 1928.
Buchthal, Hugo.
62. Early Islamic miniatures from Baghdad. In Walters Art Gal-
lery, Journal, vol. 5, pp. 19-39. Baltimore, 1942.
Buchtal, Hugo, Kurz, Otto, and Ettinghausen, Richard.
63. Supplementary notes to K. Holter's check list of Islamic il-
luminated manuscripts before A.D. 1350. In Ars Islamica,

vol. 7, pp. 147-164. Ann Arbor, Mich., 1940.
Bunt, Cyril G. E.
64. The lion and the unicorn. In Antiquity, vol. 4, pp. 425-437.
Gloucester, England, 1930.
BuzuRG IBN Shahriyar, al-Ram-R urmuzl.
65. The book of the marvels of India. From the Arabic by
L. Marcel Devic [and] translated into English by Peter
Quennell. London, G. Routledge & Sons, Ltd., 1928.
Caesar, C. Julius.
66. The Gallic war, with an English translation by H. J. Edwards
. . . London, W. Heinemann; New York, G. P. Putnam's
Sons, 1939. (Loeb Classical Library. Latin Authors.)
Cairo. Musee National Arabe.
67. Catalogue general du Musee arabe du Caire; objets en cuivre,
par Gaston Wiet. Le Caire, Impr. de l'lnstitut Frangais
d'Archeologie Orientale, 1932.
Cammann, Schuyler.
68. The development of the Mandarin square. In Harvard Journal
of Asiatic Studies, vol. 8, pp. 71- 1 30. Cambridge, Mass.,
1944.
Casanowicz, I. M.
69. Unicorn. In The Jewish Encyclopedia, ed. Isidore Singer. New
York and London, Funk and Wagnalls [c. 1916], vol. 12,
P. 344-
Charpentier, Jarl.
70. Poison-detecting birds. In London University, School of Orien-
tal and African Studies, Bulletin, vol. 5, pp. 233-242, 1928-30.
Chau Ju-Kua.
71. Chau Ju-Kua: his work on the Chinese and Arab trade in the
twelfth and thirteenth centuries, entitled Chu-fan-chi.

Tr.
from the Chinese and annotated by Friedrich Hirth and
W. W. Rockhill. St. Petersburg, Printing Office of the
Imperial Academy of Sciences, 191 1.
Ch'in ting hsi citing ku chien.

72. Supplementary catalogue of ancient bronzes belonging to the
Manchu dynasty. Supplement of the Hsi cKing ku chien
MiUl&Sls Compiled under the direction of Princes of the
First Order Chuang Yin-lu ^jtftvjz and Kuo Hung-yen
Jfc3lfi t n -P-] Chi ch'eng t'u shu kung ssu, 1905. 24 pen.
First published by order of Ch'ien Lung in 1749.
Cleveland Museum of Art.

73. Islamic art; selected examples from the loan exhibition of
Islamic art in the Cleveland Museum of Art. Cleveland,
1944.
Cohn, Carl.
74. Zur literarischen Geschichte des Einhorns. Berlin, R. Gaert-
ners Verlagsbuchhandlung, 1896-97. (Wissensch. Beilage zum
Jahresbericht der elften stadtischen Realschule zu Berlin-
Osten, 1896.)
COOMARASWAMY, ANANDA KENTISH.

75. Indian drawings; 2d series, chiefly Rajput . . . [London,
printed for the India Society at the Old Bourne Press,
1912.]
Cosmas Indicopleustes.
76. Koafxd aiyviTTLov lAovaypv Xpivriaviicq Toiroypa<f>ua. The Christian

topography of Cosmas, an Egyptian monk. Tr. from the
Greek, and ed., with notes and introduction, by J. W.
McCrindle . . . London, printed for the Hakluyt Society,
1897. (Works issued by the Hakluyt Society, No. 97.)

al-Damiri, Muhammad b. Musa.
77. Haydt al-hayawdn al-kubrd <S j& jl^-l \-=- 2 vols. Cairo, \
1319 H.
78. Al-Damiri' s Haydt al-hayawdn. A zoological lexicon. Tr. by
A. S. G. Jayakar. Vols. 1-2 1 . London, Bombay, 1906-08.
De Young, M. H., Memorial Museum, San Francisco.
79. Exhibition of Islamic art February 24 to March
. . . 22,
1937. [San Francisco, 1937.]
Dilley, Arthur Urbane.
80. Oriental rugs and carpets; a comprehensive study. New York,
London, C. Scribner's Sons, 1931.
Dimand, Maurice Sven.
81. A handbook of Muhammadan art. 2d ed., rev. and enl. New
York,The Metropolitan Museum of Art, 1944.
82. Loan exhibition of ceramic art of the Near East. New York,
Metropolitan Museum of Art, 193 1.
83. A silver inlaid bronze canteen with Christian subjects. [FGA

41.10.] In Ars Islamica, vol. 1, pp. 17-21. Ann Arbor,
Mich., 1934.
84. Unpublished metalwork of the Rasfdid sultafis of Yemen. In
Metropolitan Museum Studies, vol. 3, pt. 2, pp. 229-237.
New York, June, 1931.
al-Dimashqi, Muhammad b. Abi Talib, Shams al-din.
85. Manuel de la cosmographie du rnoyen age, traduit de l'Arabe
"Nokhbet ed-dahr fi 'Adjaib-il-birr wal-bah'r" de Shems
ed-Din Abou-'Abdallah Moh'ammad de Damas, et accom-
pagne d'eclaircissements par M. A. F. Mehren. Copenhagen,
C. A. Reitzel [etc., etc.], 1874.
Diodorus Siculus.
86. Diodorus of Sicily. With an English translation by C. H.
Oldfather. 4 vols. London, W. Heinemann, Ltd.; New
York, G. Putnam's Sons, 1933-39. (Loeb Classical Library.

Greek Authors.)
Dodgson, Campbell.
87. The story of Diirers ganda. In Fowler, Alfred, The romance
of fine prints, pp. 45-56. Kansas City, The Print Society,

1938.
Dozy, Reinhart Pieter Arne.
88. Supplement aux die tionn aires arahes. 2e. ed. 2 vols. Leide,
E. J. Brill [etc., etc.], 1927.
Druce, George C.
89. Notes on the history of the heraldic jail or yale. In The Archae-
ological Journal, vol. 68, pp. 173-199. London, 191 1.
Eberhard, Wolfram.
90. Lokalkulturen im alt en China. Erster Teil. Die Lokalkulturen
des Nordens und Western. Leiden, E. J. Brill, 1942.
(T'oung pao. Suppl. to t. 37, 1942.)
Ecochard, Michel.
91. Notes d archcologie musulmane . 1. Stereotomie de deux portails
du XII e siecle. A. Portail de la Madrassa Moqaddamiya a
Alep. In Bulletin d' Etudes Orientales, t. 7-8, pp. 98-112.
Le Caire, Impr. de l'lnstitut Frangais, 1937-38.
Elias, Elias Antoon.

92. Elias modem dictionary : Arabic-English -QjG' j-*a*Jl <SyK&\
cJ^1 4th ed. Cairo, Elias' Modern Press, 1947.
Elliot, Henry Miers.

93. The history of India as told by its own historians. The Muham-
madan period. Ed. by J. Dowson. 8 vols. London, Triibner
and Co., 1867-77.
Engel-Gros, F.
94. Catalogue des tableaux anciens, tableaux modernes, objets d'art

et de haute curiosite, importantes tapisseries composant la
collection Engel-Gros et dont la vente aura lieu a Paris,
Gallerie Georges Petit . . . Paris, Impr. Georges, 1921.
Erh-ya. WM
95. A thesaurus of terms and phrases, with commentary by Kuo
P'o Jpjf- Nanking, Chin-ling shu chii, 1895.
EUMORFOPOULOS, GEORGE.
96. The George Eumorfopoulos collection; catalogue of the
Chinese, Corean and Persian pottery and porcelain, by R. L.
Hobson ... 6 vols. London, E. Benn, Ltd., 1925-28.
Faunthrope, J. C.
97. Jungle life in India, Burma and Nepal. In Natural History,
vol. 24, pp. 175-198. New York, 1924.
Ferrand, Gabriel.
98. Relations de voyages et textes geographiques arabes, per sans et
turks, relatifs a V Extreme-Orient du VIII 6 au XVIII e

siecles. 2 vols. Paris, Ernest Leroux, 191 3-1 4.
99. he Tuhfat al-Albab de Abu Hdmid al-Andalusi al-Garndti
edite d'apres les Mss. 2167, 2168, 2170 de la Bibliotheque
Nationale et le Ms. d' Alger. In Journal Asiatique, t. 207,
pp. 1-148 and pp. 193-304. Paris, 1925.
Ferrandis, Jose.
100. Mar files drabes de Occidente ... 2 vols. Madrid, Impr. de
Estanislao Maestre, 1935, 1940. (Spain. Cuerpo Facultativo
de Archiveros, Bibliotecarios y Arqueologos.)
Field, Henry, and Prostov, Eugene.
101. Excavations at Khwarazm, 1937-1938. In Ars Islamica, vol. 6,
pp. 158-166. Ann Arbor, Mich., 1939.

Firdawsi.
102. Le livre des rois, par Abou'lkasim Firdausi, publie, traduit et
commente par m. Jules Mohl ... 7 vols. Paris, Impr.
Royale, 1838-78. (Collection Orientale.)
103. The Shdhndma of Firdausi, done into English by Arthur George
Warner . . . and Edmond Warner ... 9 vols. London,
K. Paul, Trench, Triibner & Co., Ltd., 1905-25. (Triibner's
Oriental Series.)
FONTURA DA COSTA, ABEL.
104. Les deambulations du rhinoceros de Modofar roi de Cambaye,
de 1514 d 1516. Lisbon, 1937.
Frankfort, Henri.
104a. Cylinder seals; a documentary essay on the art and religion of
the ancient Near East. London, Macmillan and Co., 1939.
Frazer, Sir James George.
105. The golden bough; a study in magic and religion. (Abridged
ed.) London, Macmillan and Co., Ltd., 1925.
Garrett, Robert.
106. Descriptive catalog of the Garrett collection of Persian, Turk-
ish and Indie manuscripts . . . in the Princeton University
library, by Mohamad E. Moghadam and Yahya Armajani,
under the supervision of Philip K. Hitti. Princeton, Prince-
ton University Press; London, H. Milford, Oxford Uni-

versity Press, 1939. (Princeton Oriental Texts, vol. 6.)
Geiss, Albert.
107. Histoire de I 'imprimerie en Egypte. In Institut d'Egypte,
Cairo, Bulletin, ser. 5, t. 1, pp. 133-157, 1907.
Giuzal'ian, L. T. (Tio3ajib5m, JI. T.) 1 D'iakonov, M. M. (ZUhkohob,
M. M.)
108. Iranskie miniattury v rukopisiakh Shakh-name. Leningradskikh
sobranii MpaHCKHe MHHHanopbi b pyKoniicax LUax-HaMS JieHHH-
rpajicKHX co6paHHft. Moscow, Leningrad, Gosudarstvennyi
Ermitazh, 1935.
109. Rukopisi Shakh-name v Leningradskikh sobranhakh PyKonncH
LHax-HaMS b JienHHrpa^CKHx co6paHHx. Leningrad, Akademiia
Nauk SSSR, 1934.
Godbey, Allen H.
109a. The unicorn in the Old Testament. In The American Jour-
nal of Semitic Languages and Literatures, vol. 56, pp. 256-
296. Chicago, 1939.
GOBEL, HEINRICH.
no. JFandteppiche ... 6 vols. Leipzig, Klinkhardt & Biermann,
1923-24.
Goeje, Michael Jan de.
in. Indices, glossarium et addenda et emendanda. Lugdunum Ba-
tavorum (Leiden), 1879. (Bibliotheca Geographorum
Arabicorum, vol. 4.)
Gotze, Alfred.
112. Friihneuhochdeutsches Lesebuch. 2. verm. aufl. Gottingen, Van-
denhoeck & Ruprecht, 1925.
GOLDSCHMIDT, AdOLPH.
113. Die Elfenbeinskulpturen unter Mitwirkung von P. L.
. . .
Hiibner und O. Homburger. 4 vols. Berlin, Cassirer,

1914-26.
Grimm, Jakob Ludwig Karl, and Grimm, Wilhelm Karl.
114. Kinder- und Hausmarchen, gesammelt durch die Briider Grimm.
3 vols. Berlin, Propylaen-verlag, 19 10.
Guest, Grace Dunham.

115. Shiraz painting in the sixteenth century. Washington, D. C,
Smithsonian Institution, 1949. (Freer Gallery of Art, Ori-
ental Studies, No. 4.)
GUILLAUME LE CLERC.
116. Le Bestiaire. Das Thierbuch des normannischen Dichters Guil-
laume le Clerc zum ersten Male vollstandig nach den Hand-
schriften von London, Paris, und Berlin mit Einleitung und
Glossar herausgegeben von . . . Robert Reinsch. Leipzig,
O. R. Reisland, 1892. ( Altfranzosische Bibliothek, Bd. 14.)
Harari, Ralph.
117. Metalwork after the early Islamic period. In Pope, A. U.,
Survey of Persian art, vol. 3, pp. 2466-2529 ; vol. 6, pis.
1 276- 1 396. London and New York, Oxford University

Press, 1939.
Harper, Francis.
118. Extinct and vanishing mammals of the old world. Illustrated
by Earl L. Poole. New York, American Committee for
International Wild Life Protection, 1945. (Special publ.,
No. 12.)
Heckscher, William S.
119. Bernini s elephant and obelisk. In The Art Bulletin, vol. 29,
pp. 155-182. New York, 1947.

Heller, Edmund.
120. The white rhinoceros. Washington, D. C, Smithsonian Institu-
tion, 1913. (Smithsonian Miscellaneous Collections, vol. 61,
No. 1.)
Hennig, R.
121. Der mittelalterliche arabische Handelsverkehr in Osteuropa.
In Der Islam, Bd. 22, pp. 239-265. Berlin, 1935.
Herzfeld, Ernst Emil.

122. Iran in the ancient East. London, New York, Oxford Univer-
sity Press, 1 94 1.
123. Die Kunst des zweiten J ahrtausends in V orderasien. In Archaeo-
logische Mitteilungen aus Iran, Bd. 8, Hft. 3, pp. 103-160,
December 1937; Bd. 9, Hft. 1, pp. 1-79, January 1938.
Berlin.
Hobson, Robert Lockhart.

124. The later ceramic wares of China . . . London, Ernest Benn,
Limited, 1925.
Holter, Kurt.
125. Die friihmamlukische Miniaturenmalerei. In Die graphischen
Kiinste, n. f., Bd. 2, Hft. 1, pp. 1-14. Wien, 1937.
126. Die islamischen Miniaturhandschriften vor 1350 Leipzig, . . .
Otto Harrassowitz, 1937. Sonderabdruck aus dem Zentral-

blatt fur Bibliothekswesen, Jahrg. 54, Hft. 1/2, pp. 1-34.
Hopkins, E. Washburn.
127. Epic mythology. Strassburg, Verlag von Karl J. Triibner, 1915.
(G. Biihler. Grundriss der Indo-arischen Philologie und
Altertumskunde 3, Bd. I, Hft. B.)
Hubback, Theodore.
128. The Asiatic two-horned rhinoceros. In Journal of Mammalogy,
vol. 20, pp. 1-20. Baltimore, 1939.
Hudud al-Alam.
129. 'The regions of the world,' a Persian geography, 3J2 A.H.-
982 A.D. Tr. and explained by V. Minorsky London, . . .
Luzac & Co., 1937. (E. J. W. Gibb Memorial Series,
n. s., vol. 11.)
Ibn Abl Usaibi'a.

130. 'Vyiin al-anba fi tabaqat al-atibba O^J (3 6
VV1
*U1*Y! ed. August Muller. 2 vols. Cairo, A.D. 1882, and
Konigsberg i. Pr., Selbstverlag, 1884.
Ibn al-Balkhi.
131. Description of the province of Fars in Persia at the beginning
of the twelfth century A.D. Tr. by Guy Le Strange. In
Royal Asiatic Society of Great Britain and Ireland, Journal,
pp. 1-30, London, 1912.
Ibn al-Balkhi.
132. The Fdrsndma of Ibnu 'l-Balkhi. Ed. by G. Le Strange and
R. A. Nicholson. London, Luzac & Co., 192 1. (E. J. W.
Gibb Memorial Series, n. s., vol. 1.)
Ibn Battuta.
I 33- Voyages d'lbn Batoutah. Texte arabe, accompagne d'une tra-
duction par C. Defremery et le Dr. B. R. Sanguinetti. 4 vols.
& index vol. Paris, Impr. par autorisation du Gouvernement

a rimprimerie Nationale, 1914-1927.
Ibn Fadlan.
134. Ibn Fadlan s Reisebericht, von A. Zeki Validi Togan . . . Leip-
zig, Kommissionsverlag F. A. Brockhaus, 1939. (Deutsche
Morgenlandische Gesellschaft, Abhandlungen fur die Kunde
des Morgenlandes, XXIV, 3.)
Ibn Hawqal.
135- Viae et regno, descriptio ditionis moslemicae, auctore Abul-
Kdsim Ibn HaukaL uOUJlj Ed. M. J. de
Goeje. Lugdunum Batavorum (Leiden), E. J. Brill, 1873.

(Bibliotheca Geographorum Arabicorum, vol. 2.)
Ibn Khordadhbeh.
136. Kitab al-masalik wa 'l-mamalik {liber viarum et regnorum)
auctore Abu'l Kasim Obaidallah ibn Abdallah Ibn Khordadh-
beh et excerpte e Kitab al-Kharadj auctore Kodama ibn
Djafar. c-OUJlj cÔUil Jtf Ed. M. J. de Goeje. Lug-

dunum Batavorum (Leiden), E. J. Brill, 1889. (Bibliotheca
Geographorum Arabicorum, vol. 6.)
India. Archaeological Survey.

137. Excavations at Harappa, 1920-31 and 1933-34, by Madho
Sarup Vats ... 2 vols. Delhi, Manager of Publications,
1940.
Isidorus, bp. of Seville.
138. Etymologiae. In Migne, Jacques Paul. Patrologiae cursus com-
putus . . . series latina, vol. 82. Paris, Gamier Freres,
1878.
al-IstakhrT.
139. Viae regnorum descriptio ditionis moslemicae auctore Abu Ishak
al-Fdrisi al-Istakhri. uOUJl lJOL^ Ed. M. J.
de Goeje. Lugdunum Batavorum (Leiden), E. J. Brill,
1870. (Bibliotheca Geographorum Arabicorum, vol. 1.)
Ivins, William Mills.

140. A neglected aspect of early print-making. In Metropolitan
Museum of Art, Bulletin, vol. 7, No. 2, p. 51-59. New York,
October 1948.
Jacob, Georg.
141. Kleinere Mitteilungen zum V erstdndnis Qazwlnis. In his
Studien in arabischen Geographen, Hft. 4, pp. 1 51-176. Ber-

lin, 1892.
142. Die JVaaren beim arabisch-nordischen Verkehr im Mittelalter
. . . Berlin, Mayer & Muller, 1891.
143. Welche Handelsartikel bezogen die Araber des Mittelalters aus
den nordisch-baltischen Ldndern ... 2. aufl. Berlin, Mayer
& Muller, 1891.
Jahangir, Emperor of Hindustan.

144. The Tuzuk-i-Jahdngirl, or, Memoirs of Jahangir Tr. by . . .
Alexander Rogers Ed. by Henry Beveridge ... 2 vols.

. . .
London, Royal Asiatic Society, 1909-14. (Oriental Transla-

tion fund, n. s.j vols. 19, 22.)
al-Jahiz.
145. Kitab al-hayawan jl^-l 7 vols. Cairo, 1323-1325 H.
(A.D. 1905-1907).
146. Kitab al-hayawan b\$^*\ Edited by 'Abd al-Salam
Muhammad Harun. 7 vols. Cairo, Mustafa al-Babi al-
Halabi & Sons, 1938-45.
Japan. Imperial Treasury (Shosoin).
147. Shosoin Gyomotsu zuroku JEjti%$P%[S$ Illustrated cata-
logue of the Imperial treasures in the Shosoin, vols. 1-12.
Tokyo, Imperial Household Museum, 1929-40. In Japanese

with English notes.
Jerdon, T. C.
148. The mammals of India; a natural history of all the animals
known to inhabit continental India. London, John Wheldon,
1874.
Johnson, Francis.
149. A dictionary ; Persian, Arabic and English . . . London, W. H.
Allen and Co., 1852.
Kahle, Paul.
150. Die Schatze der Fatimiden. In Deutsche Morgenlandische
Gesellschaft, Zeitschrift, Bd. 89, n.f. 14, pp. 329-362.
Leipzig, 1935.
Knighton, William.
151. The private life of an eastern king ... Ed. with an introduc-
tion and notes by S. B. Smith. London, New York, H. Mil-
ford, Oxford University Press, 1921.
Kramers, J. H.
152. Djughrdfiyd. In Encyclopaedia of Islam. Suppl. vol., pp. 61-73.
Leiden, E. J. Brill; London, Luzac & Co., 1938.
Kraus, Paul.
153. Jdbir ibn Hayydn, contribution a Vhistoire des idees scientifigues
dans VIslam, t. 1-2. Cairo, 1942-43. (Institute d'Egypte,
Cairo. Memoires, t. 44-45.)
Kuhnel, Ernst.
154. History of miniature painting and drawing. In Pope, A. U.,
Survey of Persian art, vol. 3, pp. 1 829- 1897 vol. 5, pis. 812- >'
925. London and New York, Oxford University Press,

1938-39.
155. Holz und Elfenbein. In Sarre, F., and Martin, F. R. Die
Ausstellung von Meisterwerken muhammedanischer Kunst
in Munchen, 1910, Bd. 3, pp. 1-3, Tafeln 246-257.
Miinchen, F. Bruckmann a.-g., 19 12.
156. Jagdbilder aus Indien. In Atlantis, Jahrg. 13, pp. 417-424.
Berlin-Grunewald, 1941.
157. Das Q azwini-fragment der Islamischen Abteilung. In Jahrbuch
der Preussischen Kunst sammlun gen, Bd. 64, pp. 59-72. Ber-
lin, I943-
Kushi Takushin Ûj^^Jt
158. Shina Min-sho toji zukan ^MW^MWl H Tokyo, Houn
sha, 1943.
Lamm, Carl Johan, and Geiger, Agnes.
159. En Miniatyr och en matt a; en studie over sentimuridiska gro-
teskmattor frdn Herat ... In Stockholm, Nationalmusei
Arsbok, n.s. I, pp. 26-51. 1930-31.
Lamm, Carl Johan.
160. Mittelalterliche Glaser und Steinschnittarbeiten aus dem Nahen
Osten . . . 2 vols. Berlin, D. Reimer, 1930, '29. (Forschun-
gen zur Islamischen Kunst . . . Bd. 5.)
Lane, Frank Walter.

161. Nature parade. 3d ed. London, Jarrolds, 1946.
Lang, Herbert.
162. The white rhinoceros of the Belgian Congo. In New York
Zoological Society, Bulletin, vol. 23, pp. 66-92. 1920.
Lauchert, Friedrich.
163. Geschichte des griechischen Physiologus. Strassburg, 1889.
Laufer, Berthold.
164. Arabic and Chinese trade in walrus and narwhal ivory. In
T'oung pao, ser. 2, t. 14, pt. I, pp. 315-364. Leiden, 1913.
Addenda par Paul Pelliot, pp. 365-370.
Supplementary notes on walrus and narivhal ivory.
In T'oung pao, ser. 2, t. 17, pp. 348-389. Leiden, 1916.
165. History of the rhinoceros. In his Chinese clay figures . . . pp.
73-173. Chicago, 1914. (Field Museum of Natural His-
tory, Publ. 177, Anthropological Series, vol. 13, No. 2.)
166. Sino-lranica ; Chinese contributions to the history of civilization
in ancient Iran. Chicago, 1919. (Field Museum of Natural
History, Publ. 201, Anthropological Series, vol. 15, No. 3.)
4
London. International Exhibition of Persian Art, 193 i.

167. Persian art, an illustrated souvenir of the exhibition of Persian
art at Burlington House, London, 1 931. 2d ed. [London],
Hudson & Kearns, Ltd., 1931.
Lucas, Frederic A.
168. The unicorn and his horn. In Natural History, vol. 20, pp. 533-
535. New York, 1920.
LUDERS, HEINRICH.
169. Die Sage von Rsyasrhga. In Gesellschaft der Wissenschaften
zu Gottingen. Phil.-hist. klasse. Nachrichten, 1897, PP 87-
135; 1901, pp. 28-56.
Lydekker, Richard.
170. The game animals of India, Burma, Malaya and Tibet; being
a new and revised edition of The great & small game of
India, Burma, & Tibet. London, Rowland Ward, Ltd.,
1907.
Mackay, Ernest John Henry.
171. Chanhu-daro excavations, 1935-36 New Haven, Conn.,
. . .
Publ. for American School of Indie and Iranian Studies and

Museum of Fine Arts, Boston, by American Oriental Society,
1943. (American Oriental Series, vol. 20.)
MACY, V. EVERIT.
172. The V. Everit Macy collection including rare and important
Persian and Mesopotamian pottery, Persian and Indian minia-
tures, Persian brocades and velvet carpets . . . to be sold at
auction . . . January sixth, seventh, eighth at two o'clock.

New York, American Art Association, 1938. (Sale number
4360.)
Mahmud al-Kashghari.
173. Diwanlughdt al-Turk ^J^\ d^Jt^ 3 vols. Istanbul,
Maarif Nezareti, 1917-1919.
al-Ma'luf, Amin Fahd.
174. Mujam al-hayawdn b\5r~\
f?**
An Arabic zoological dic-
tionary. Cairo, al-Muqtataf Press, 1932.

al-Maqdisi (al-Muqaddasi) .
175. Descriptio imperii moslemici. ^^V^ j** <3 Ed.

M. J. de Goeje. Lugdunum Batavorum (Leiden), 1877.
(Bibliotheca Geographorum Arabicorum, vol. 3.)
Marcais, Georges.
176. Les figures d'hommes et de betes dans les bois sculptes d'epoque
fdtimide conserves au Musee arabe du Caire. Etudes d'icono-
graphique musulmane. In Melanges Maspero, t. 3. Orient

islamiqiie. Le Caire, Impr. de l'lnstitut Frangais d'Archeo-
logie Orientale, 1940, pp. 241-257. (Cairo, Institut Fran-
cais d'Archeologie Orientale, Memoires, t. 68.)
Markwart, Josef.
177. Ein arabischer Bericht iiber die arktischen (uralischen) Lander
aus dem 10. Jahrhundert. In Ungarische J ahrbucher } Bd. 4,
pp. 261-334. Berlin, 1924.
178. Osteuropaische und ostasiatische Streifziige . . . Leipzig,

Dieterich'sche Verlagsbuchhandlung, Theodor Weicher, 1903.
Marshall, Sir John Hubert.
179. Mohenjo-daro and the Indus civilization ... 3 vols. London,
A. Probsthain, 1931.
Marteau, Georges, et Vever, Henri.
180. Miniatures persanes tirees des collections de mm. Henry d' Alle-
magne, Claude A net [e.a~\ . . . et exposees au Musee des
arts decoratifs juin-octobre igi2. Preface et commentaire par
Georges Marteau et Henri V ever. 2 vols. Paris, Bibliotheque
d'Art et Archeologie, 19 13.
Martin, Frederik Robert.

181. Altere Kupferarbeiten aus dem Orient. Gesammelt von F. R.
Martin. Stockholm, 1902.
182. The miniature painting and painters of Persia, India and
Turkey from the 8th to the 18th century ... 2 vols. Lon-
don, Bernard Quaritch, 191 2.
Marvazi.
183. Sharaf al-Zamdn Tdhir Marvazi on China, the Turks and
India. With an English translation and commentary by
V. Minorsky. London, The Royal Asiatic Society, 1942.
(James G. Forlong Fund, vol. 22.)
al-Mas'udi.
184. Les prairies d'or. Texte et traduction par C. Barbier de
Meynard et Pavet de Courteille. 9 vols. Paris, Impr. Im-
periale, 1861-1917.
Migeon, Gaston.
185. Exposition des arts musulmans au Musee des arts decoratifs.
Paris, Librairie Centrale des Beaux-arts, 1903.
186. Mussulman art [The Louvre Museum]. 2 vols. Paris, Albert
Morance, c. 1922.
13
Moeller, Marian Estabrook.

187. An Indo-Portuguese embroidery from Goa. In Gazette des
Beaux-arts, 6th ser., vol. 34, pp. 1 17-132. Paris, 1948.
Monier- Williams, Sir Monier.
188. A Sanskrit-English dictionary, etymologically and philologically
arranged with special reference to cognate Indo-European
languages . . . New ed., greatly enl. and improved. Oxford,
The Clarendon Press, 1899.
Moser, Henri.
189. Collection Henri Moser-Charlottenf els . . . Oriental arms and
armour . . . Leipzig, K. W. Hiersemann, 1912.
Muller, August.
190. Des spanisch-arabischen Reisenden Abu Hamid Kosmographie
"Tuhfat al-albab" und Hire wissenschaftliche Ausbeute. In
Jacob, Georg, Studien in arabischen Geographen, Hft. 3,
pp. 69-94. Berlin, 1892.
al-Mustawfi, al-Qazwini, Hamd Allah.

191. The zoological section of the N uzhatu-l-qulub of Hamdullah
al-Mustaufi al-Qazwini. Ed., tr., and annotated by Liejit.-
Col. J. Stephenson . . . London, The Royal Asiatic Society,
1928. (Oriental Translation Fund, n.s., vol. 30.)

al-Nadim.
192. Kitab al-Fihrist C~>j$\ o^. Mit Anmerkungen herausgege-
ben von Gustav Flugel. Nach dessen Tode besorgt von Dr.
Johannes Roediger und Dr. August Muller. 2 vols. Leipzig,
F. C. W. Vogel, 1871-72.
New York. Metropolitan Museum of Art. Cochran Collec-
tion.
193. A catalogue of the collection of Persian manuscripts including
also some Turkish and Arabic, presented . . . by Alexander
Smith Cochran. Prepared and edited by A. V. Williams
Jackson and Abraham Yohannan. New York, Columbia
University Press, 1 91 4. ( Columbia University. Indo-Iranian
Series, vol. 1.)
New York. Metropolitan Museum of Art.

194. A guide to an exhibition of Islamic miniature painting and book
illumination, by M. S. Dimand. New York, October 9
through January 7, 1933-1934. [Portland, Me., The South-
worth Press, 1933.]
195. A handbook of the loan exhibition of French tapestries, mediae-

val,Renaissance and modern November 21, IQ'47 to
. . .
February 2Q, IQ48. New York, 1947-

196. The unicorn tapestries at the Cloisters: a picture book. New
York, 1946.
NlLAKANTHA.
197. The elephant-lore of the Hindus; the elephant-sport (Matanga-
lila) of Nilakantha. Tr. and explained by Franklin Edgerton.
New Haven, Yale University Press, 1931.
Nizamu 'd-din, Muhammad.

198. Introduction to the Jawdmi ul-hikdydt wa lawdmi ur-riwdydt
of Sadidu 'd-din Muhammad al-Awfi . . . London, Luzac
& Co., 1929. (E. J. W. Gibb Memorial Series, n.s., vol. 8.)
Noldeke, Theodor.
199. Beitrage zur Geschichte des Alexanderromans. In Akademie
der Wissenschaften. Phil-hist. klasse, Denkschriften, Bd. 83,
Abh. 5. Wien, 1890.
200. Das iranische Nationalepos ... 2. aufl. Berlin, Leipzig, Verei-
nigung Wissenschaftlicher Verleger, 1920.
al-Nuwairi.
201. Nihdyat al-arab fi funun al-adab o^Y! Qy$ <3 ojVI *V
13 vols. Cairo, Dar al-kutub al-Misriya, 1342-1357 H./
A.D. 1923-1938.
Oppian.
202. Oppian, Colluthus, Tryphiodorus, with an English translation
by A. W. Mair. London, W. Heinemann ; New York, G. P.
Putnam's Sons, 1928. (Loeb Classical Library. Greek
Authors.)
Paris. Bibliotheque Nationale. Dept. des Manuscrits.
203. Catalogue des manuscrits arabes. Par M. le Baron de Slane
. . . Paris, Impr. Nationale, 1883-1895.
Peters, John P. [etc.]
204. Fainted tombs in the necropolis of Marissa. London, Palestine
Exploration Fund, 1905.
Philostratus, Flavius.
205. A pollonius of Tyana, the epistle of Apollonius and the treatise
of Eusebius. With English translation by F. C. Conybeare.
2 vols. London, William Heinemann ; New York, G. P.
Putnam's Sons, 1927. (Loeb Classical Library. Greek
Authors.)
Photius, Patriarch of Constantinople.

206. Myriobiblon sive bibliotheca. In Migne, Jacques Paul,
Patrologiae cursus completus . . . series graeca, t. 104.
Paris, Garnier Freres, 1896.
Physiologus.
207. Das "Buch der N aturge genstande " herausgegeben und iibersetzt
von Karl Ahrens . . . Kiel, C. F. Haeseler, 1892.
Physiologus.
208. Ed. by F. Sbordone. Mediolani [etc.] in sedibus societates
"Dante Alighieri-Albrighi, Segati, etc.," 1936.
Pliny the Elder.

209. Natural history. With an English translation by H. Rackham
... 4 vols. Cambridge, Mass., Harvard University Press;
London, W. Heinemann, Ltd., 1938-45. (Loeb Classical
Library. Latin Authors.)
Polo, Marco.
210. The book of Ser Marco Polo. Tr. and ed. by Sir Henry Yule.
3d ed. revised by Henri Cordier. 2 vols. London, John
Murray, 1921.
211. The travels of Marco Polo. Tr. and ed. by William Marsden,
re-edited by Thomas Wright. London, J. M. Dent & Sons,
Ltd.; New York, E. P. Dutton & Co., Inc., 1946. (Every-
man's Library, No. 306.)
Pope, Arthur Upham.
212. The art of carpet making. In his Survey of Persian art, vol. 3,
pp. 2257-2430; vol. 6, pis. 1 107-1275. London and New

York, Oxford University Press, 1939.
213. The ceramic art in Islamic times; a history. In his Survey of
Persian art, vol. 1, pp. 1446-1664; vol. 5, pis. 555-911. Lon-
don and New York, Oxford University Press, 1938.
Prisse d'Avennes, Achille Constant Theodore Emile.
e
214. Uart arabe d'apres les monuments du Kaire depuis le vii siecle
e
jusqua la fin du xviii . . . text and 3 vols. Paris, A. Morel
et C ie
, 1877.
Pseudo-Callisthenes.
215. The history of Alexander the Great, being the Syriac version
of the Pseudo-Callisthenes. Ed. from five manuscripts with
an English translation and notes by E. A. W. Budge . . .
Cambridge, University Press, 1889.

al-Qazwini, Zakariya' b. Muhammad.

216. Die Abschnitte uber die Geister und wunderbaren Geschopfe aus
Oazwinis Kosmographie iibertragen und mit Anmer-
. . .
kimgen versehen von Jonas Ansbacher. Kirchhain N.-L.,

M. Schmersow, 1905. (Inaug.-diss., Erlangen.)
217. Zakariya ben Muhammed ben Mahmud el-Cazwinis Kosmo-

graphie. Erster Theil: Die W under der Schopfung. Ed.
by F. Wiistenfeld. 2 vols. Gottingen, Dieterichsche Buch-
handlung, 1849.
218. Zakarija ben Muhammed ben Mahmud el-Kazwini's Kosmo-
graphie . . . iibersetzt von H. Ethe. Halbbd. 1. Die W un-
der der Schopfung. Leipzig, Fues, 1868.
Reinhart, Werner.
219. Zur Identifizierung von "al-Chutww." In Der Islam, Bd. 3,
p. 184. Strassburg, 1912.

Robinson, Theodore Henry, Hunkin, Joseph W., and Burkitt,
Francis C.
220. Palestine in general history. London, published for the Brit-
ish Academy by Humphrey Milford, 1929. (The Schweich
Lectures of the British Academy, 1926.)
Ross, Marvin Chauncey.
221. Eine deutsche W alroszschnitzerei romanischer Zeit. In Pan-
theon, Bd. 24, pp. 243-244. Munich, 1939.
Ruska, Julius.
222. Noch einmal al-Chutww. In Der Islam, Bd. 4, pp. 163-1640
Strassburg, 1913.
Russia. Arkheologicheskaia Kommissiia.
223. BocTOHHoe cepe6po. Auiacb ApeBHeft cepe6pfmoii h 30,noToft nocy/ibi
BOCTOHHaro npoHcxo>K,neHiH, HaftêHHOH npeHMymecTBeHHO btj
npejrtjiax'b PoccificKOH Hjvmepin. H3,naHie MMnepaiopcKOH apxeo-
jiorHêcKofi KOMMHCciii ko ahk) nHTHAecHTHJTBTiH ea ;rtHTe-
jibHOCTH. C.-rieTepSypr-b, 1909.
Sabri Sharif pasha.
224. Miniatures persanes, iurques et indiennes; collection de son
excellence Cherif Sabry pacha. [Avant-propos par] Gaston
Wiet. [Le Caire, Impr. de 1'Institut Francais d'Archeologie
Orientale, 1943] (Memoires Presentes a 1'Institut d'Egypte

. . . t. 47.)
Sarre, Friedrich Paul Theodor.
225. Erzeugnisse islamischer Kunst, Teil II. Seldschukische Klein-
kunst (Sarre collection) . Leipzig, 1909.
226. Die Erwerbung einer in Sudrussland gebildeten Sammlung von

Werken der Kleinkunst aus antiker, Volkerwanderungs- und
islamischer Zeit. In Berliner Museen; Amtliche Berichte aus
den preussischen Kunstsammlungen, Bd. 29, pp. 65-71. Ber-
lin, 1907/08.
227. Der Kiosk von Konia . . . Berlin, Verlag fur Kunstwissen-

schaft, 1936.
228. Die Kunst des alten Persian Berlin, Bruno Cassirer, 1922.
. . .
(Die Kunst des Ostens Bd. 5.) . . .
229. Ein neuerworbenes Beispiel mesopotamischer Relief keramik. In

Berliner Museen; Amtliche Berichte aus den preussischen
Kunstsammlungen, Bd. 51, pp. 7-1 1. Berlin, 1930.
230. Sammlung F. u. M. Sarre. Katalog der Ausstellung im Sta-
delschen Kunstinstitut. Frankfurt am Main, 1932.
231. Die Tradition in der iranischen Kunst. In International Con-
gress for Iranian Art and Archaeology, 3rd, Leningrad and
Moscow, 1935. Memoirs, pp. 227-228, pis. 103- 106. Moscow
& Leningrad, Akademiia Nauk SSSR, 1939.
Sarre, Friedrich Paul Theodor, and Herzfeld, Ernst.
232. Iranische Aufnahmen und Untersuchungen von
Felsreliefs ;
Denkmalern aus alt- und mittelpersischer Zeit Berlin, . . .
E. Wasmuth, a.-g., 1910.

Sarre, Friedrich Paul Theodor, and Martin, F. R.
233. Die Ausstellung von Meisterwerken muhammedanischer Kunst
in M'unchen, iqio ... 3 vols. Miinchen, F. Bruckmann,
a.-g., 1912.
Sarton, George.
234. Introduction to the history of science ... 3 vols, in 5. Balti-
more, publ. for the Carnegie Institution of Washington by
the Williams and Wilkins Company, 1927-1948. (Carnegie
Institution of Washington, Publ. No. 376.)
Sauvaget, Jean.
235. Alep; essai sur le developpement d'une grande ville syrienne,
des origines au milieu du xix e sVecle ... 2 vols. Paris, Paul

Geuthner, 1941. (Haut Commissariat de l'Etat Frangais
en Syrie et au Liban. Service des Antiquites. Bibliotheque
Archeologique et Historique, t. 36.)
Schapiro, Meyer.
236. On the aesthetic attitude in romanesque art. In Art and
Thought, pp. 130-150. Ed. by K. Bharatha Iyer. London,
Luzac & Company, Ltd., 1947.
Schmidt, J. Heinrich.
237. Persische Seidenstoffe der Seldjukenzeit . . . In Ars Islamica,
vol. 2, pp. 84-91. Ann Arbor, Mich., 1935.
SCHONBERGER, GuiDO.
238. N arwhal-Einhorn ; Studien iiber einen seltenen Werkstoff. In
Stadel-jahrbuch, Bd. 9, pp. 167-247. Frankfurt am Main,
1935/36.
Schulz, Philipp Walter.
239. Die persisch-islamische Miniaturmalerei ... 2 vols. Leipzig,
Karl W. Hiersemann, 1914.
Selous, F. C.
240. Mammals of other lands. In Nature Lovers Library, vol. 5,
p. 160. New York, The University Society, Inc., 191 7.

Shebbeare, E. O.
241. Protecting the great Indian rhinoceros. In The Field, vol. 165,
pp. 1229-1231. London, May 18, 1935.

Shepard, Odell.
242. The lore of the unicorn. London, George Allen & Unwin, Ltd.
[i930.]
Shortridge, G. C.
243. The mammals of Southwest Africa; a biological account of the
forms occurring in that region. 2 vols. London, William
Heinemann, Ltd., 1934.
Siren, Osvald.
244. Chinese paintings in American collections. 4 vols. Paris and
Brussels, G. van Oest, 1927.
Sirich, Marjorie W.
245. Gift of Persian miniatures (by the late Charles C. Webber and
Mrs. Webber). In Minneapolis Institute of Arts, Bulletin,
April 7, 1945. (Entire issue.)
Sowerby, Arthur de Carle.
246. Nature in Chinese art with two appendices on the Shang
. . .
pictographs, by Harry E. Gibson. New York, The John

Day Company [c. 1940].
Steier.
247. Nashorn. In Pauly's Real-Encyclopadie der classischen Alter-
tumswissenschaft, by A. Pauly and G. Wissowa, cols. 1780-
1788. Stuttgart, Metzler, 1894-1935.
248. Timotheus, No. 18. In Pauly's Real-Encyclopadie der classischen
Altertumswissenschaft, by A. Pauly and G. Wissowa, cols.
I339-I34I- Stuttgart, Metzler, i894"*935-

Steindorff, George.
249. The magical knives of ancient Egypt. In Walters Art Gallery,
Journal, vol. 9, pp. 41-51, 106-107. Baltimore, 1946.
Steingass, Francis Joseph.
250. A comprehensive Persian-English dictionary . . . London,
K. Paul, Trench, Trubner & Co., Ltd., 1930.
Stone, George Cameron.

251. A glossary of the construction, decoration and the use of arms
and armor in all countries and in all times. Portland, Maine,
Southworth Press, 1934.
Stott, Ken.
252. Too many lions. In Zoonooz, vol. 21, No. II, pp. 3-4. San
Diego, November 1948.
Strabo.
253. The geography of Strabo with an English translation by Horace
Leonard Jones Based in part upon the unfinished ver-
. . .
sion of John Robert Sitlington Sterrett ... 8 vols. London,

W. Heinemann; New York, G. P. Putnam's Sons, 1917-33.
(Loeb Classical Library. Greek Authors.)
Sulaiman, the Merchant. See Akhbar al-Sin wa'l-Hind.
SUTTANIPATA.
254. The Sutta-nipdta, a collection of discourses ; being one of the
canonical books of the Buddhists, translated from Pali by
V. Fausboll. 2d ed., rev. London, Humphrey Milford,
Oxford University Press, 1924. (The sacred books of the
East . . . vol. 10, pt. 2.)
al-Tawhidi, Abu Haiyan Ali b. Muhammad b. al-'Abbas.
255. Kitab al-Imta wa'l-mu dnasa A*J|J^Jli ^Y1 Edited by

Ahmad Amin and Ahmad al-Zain. Cairo, 1939-44.
Tennent, Sir J. Emerson.

256. The wild elephant and the method of capturing and taming it
in Ceylon. London, Longmans, Green and Co., 1867.

Textile Museum.
257. Dragon rugs, a loan exhibition from American public and pri-
vate institutions . . . October 18 through November IQ,
1948. Washington, D. C, 1948.

Thom, W. S.
258. Rhinoceros shooting in Burma. In Bombay Natural History

Society, Journal, vol. 38, pp. 137-150. 1936.
Timotheus Gazaeus.
259. In Haupt, Moriz, Opuscula, Bd. 3, pp. 274-302. Leipzig,
Solomon Hirzel, 1875-76.
Togan, Ahmed Zeki Validi.
260. Die Nordvolker bei Biriini. In Deutsche Morgenl'dndische
Gesellschaft, Zeitschrift, Bd. 90, n.f. 15, pp. 38-51. Leipzig,
1936.
Umehara Sueji ftlKîcjp
261. Shina-kodo seikwa '^ffl'feWiW^ or selected relics of ancient
Chinese bronzes from collections in Europe and America . . .
7 vols. Osaka, Yamanaka & Co., I933"35-
Validi, A. Zeki. See Togan, Ahmed Zeki Validi.

Van Berchem, Max.
262. Inscriptions mobilieres arabes en Russie. In Journal Asiatique,
I0 e ser., t. 14, pp. 401-413. Paris, 1909.
Vaughan-Kirby, F.
263. The white rhinoceros, with special reference to its habits in
Zululand. In Annals of the Durban Museum, vol. 2, pp. 223-
242. Durban, Natal, 1917/20.

Veselovskii, N. I. (BecejioBCKHft, H. H.)
264. Geratskii bronzovyi kotelok S59 goda gidzhry (1163 g. po

R. Khrist.) iz sobraniia grafa A. A. Bobrinskago. TepaTCKift
6poH30Bbitt KOTe.JioKT> 559 roâ rnA>Kpbi (1163 r. no P.X.) nsi.
co6paHin rpa$a A.A. Bo6pnHCKaro. St. Petersburg, Tipografiia
Imperatorskoi Akademiia Nauk, 19 10. (Materialy po
Arkheologii Rossii izd. Imp. Arkheologicheskoiu Kommis-
No. 33.)
sieiu,
Vienna. Osterreichisches Museum fur Kunst und Industrie.

265. Old oriental carpets . . . with text by Friedrich Sarre and
Hermann Trenkwald, tr. by A. F. Kendrick. 2 vols. Vienna,
A. Schroll & Co., 1926-29.
Visser, Herman Floris Eduard.
266. Asiatic art in private collections of Holland and Belgium. Am-
sterdam, "De Spieghel" Publishing Co., 1948.
Von der Osten, Hans Henning.
267. The ancient seals from the Near East. In The Art Bulletin,
vol. 13, pp. 221-241. New York, 1931.
Vullers, Johann August.
268. loannis Augusti Vullers lexicon persico-latinum etymologicum
. . . 2 vols. Bonnae ad Rhenum. impensis A. Marci, 1855-64.
i go Freer Gallery of Art Occasional Papers Vol. i
Ward, Rowland.
269. Records of big game with their distribution, characteristics,
dimensions, weights and horn & tusk measurements. 6th ed.
London, Rowland Ward Ltd., 1910.
Ward, William Hayes.

270. The seal cylinders of western Asia. Washington, D. C, Car-
negie Institution, 1910.
Warner, Sir George Frederic.

271. Queen Mary's Psalter, miniatures and drawings by an English
artist of the 14th century. Reproduced from Royal Ms.
2B-VII in the British Museum. London, 191 2.
Weber, Otto.
272. Altorientalische Siegelbilder. 2 vols. Leipzig, J. C. Hinrichs'sche
Buchhandlung, 1920. (Der alte Orient, Bd. 17 & 18.)
Weber, Victor Frederic.

273. "Ko-ji ho-ten" ; dictionnaire a Vusage des amateurs et collec-
tionneurs d'ob)ets d'art japonais et chinois ... 2 vols. Paris,

L'auteur, 1923.
Wellmann, M.
274. Elefant. In Pauly's Real-Encyclopadie der classischen Alter-
tumswissenschaft, by A. Pauly and G. Wissowa, cols. 2248-
2257. Stuttgart, Metzler, 1894-1935.
275. Der Physiologus. Eine religionsgeschichtlich-naturwissenschaft-

liche Untersuchung. In Philologus, Supplement Band, Bd.
22, p. 116. Leipzig, 1931.
West, Edward William, ed. and tr.
276. Pahlavi texts . . . Pt. I. The Bundahis, Bahman yast, and

Shayast La-shay ast. Oxford, The Clarendon Press, 1880.
(The sacred books of the East . . . vol. 5.)
Wiedemann, Eilhard.
277. Uber den Wert von Edelsteinen bei den Muslimen. In Der
Islam, Bd. 2, pp. 345-358. Strassburg, 191 1.
Williamson, Thomas.
278. Oriental field sports; being a complete, detailed, and accurate
description of the wild sports of the East; and exhibiting, in
a novel and interesting manner, the natural history of the
elephant, the rhinoceros, the tiger, etc. ... 2 vols. Draw-
ings by Samuel Howitt. London, H. R. Young, 18 19.
Wittkower, Rudolf.
279. Miraculous birds. 2. 'Roc,' an Eastern prodigy in a Dutch
engraving. In Warburg Institute, Journal, vol. I, pp. 255-
257. London, 1937-1938.
Wolff, Fritz.
280. Glossar zu Firdosis Schahname . . . Berlin, Gedruckt in der
Reichsdruckerei, 1935. 1 v l- an d supplement.
Yaqut ibn 'Abd Allah.
281. Jacut's geographisches Worterbuch aus den Handschriften zu
Berlin, St. Petersburg und Paris, auf Kosten der Deutschen
Morgenlandischen Gesellschaft herausgegeben von Ferdinand
Wiistenfeld. 6 vols. Leipzig, in Commission bei F. A.
Brockhaus, 1866- 1890.
Yule, Sir Henry, and Burnell, A. C.
282. Hobson-Jobson, a glossary of colloquial Anglo-Indian words and
phrases New ed. edited by William Crooke
. . . Lon- . . .
don, John Murray, 1903.

INDEX
Abano, Pietro d' (quoted), no.
Abu Haiyan al-Tawhidi. See al-Tawhidi, Abu Haiyan.
Abu 'l-Fa^l 'Miami. See 'Allami, Abu '1-Fadl.
Achlis, 100.
Aelian, 77, 91, 99, 113.
Aeolian flute-type horns, 64-66, 150; pi. 42.
Agatharchides, 77, 78.

Ahmad Tusi, 35, 53, 60, 102.
'Ajd'ib al-makhluqdt mss., 8, 14, 17, 65, 68, 156, 161; pis. 13-17, 22, 38, 40-44.
Al-Akfani, 114, 141.
Akhbdr al-Sin iva'l-Hind, 11, 15, 20, 53, 54, 58, 94, 99, 100.
Albertus Magnus, Saint, no.
Aleppo, sculpture from, 4, 51, 134, 135 136; pi. 6.
Alexander the Great (Iskandar), 8, 66; pis. 9, 24.
Alf laila ava-laila, n, 30, 32, 53.

Alicorn, no.
'Allami, Abu'1-Fadl, 11, 13, 17.
Allegorical interpretation of Christian-Arabic version, 92-93.
Amalgamation of khutu horn with the horn of the karkadann, 151-152.
Anatolia, 4
Andersson, Karl Johan, 132.
(quoted), 79, 80.
Animal Kingdom, 49-50; pis. 36-38.
Animal types in the Shdh-ndmah manuscripts, 43, 47; pis. 9, 23-30, 35.
Animals:
Fantastic, 1-3, 69-72.
Fights of, 26-34, 67, 78-90, 147, 160-161; fig. 2; pis. 3, 17-22, 33.
Flame-endowed, 105, 156; pi. 48.
Horns, number of, 18-23, 27, 61, 62, 64, 67, 70, 79, 108.
Hypnotized, 48.
Majnun surrounded by, 49; pi, 36.
Representations in art:
Bedspread from Goa, 97; fig. 3.
Bucket, copper-inlaid, 3, 25; pi. 4.

Candlesticks, 27.
Canteen, silver-inlaid, 3; pi. 3.
Carpets, 26, 27, 28, 32, 160; pis. 18, 21.

Friezes, 26-29; pis- 3> 19.
Ivory boxes, 4, 24, 71, 122-130, 135; pi. 5.
Manuscripts. See under names of museums and titles of manuscripts.

Mirrors, Chinese, 105, 108, 156; figs. 4-5; pi. 48.
Paintings, 10, 15, 48, 50, 59, 106, 155; pis. 7-18, 22-30, 32-38, 40-46, 48.
Plaque, 4, 35, 51; pi. 6.
193
:
Animals Continued
Representations in art Continued
Plates, silver-inlaid bronze, 26, 106, 160; pi. 18.
Sculpture, 17, 43, 67, 69, 74, 149; pi. 31.
Tiles, 26, 27; pis. 4, 18, 19.

Vessels
Bronze, 26, 27; pis. 4, 18, 19.
Glass, 3, 17, 25, 28; pis. 1, 20.
Pottery, 74, 134, 136, 161; pi. 31.
Ankle-bone of the unicorn, 61.
'Anqd', 32-33, 151; pis. 8, 17, 18.
Antelope, 18, 38, 41, 61, 71, 155; pis. 3, 29.

"Apocryphal" unicorns, 68-70; pis. 17, 45.
Arab beliefs:
Concerning the unicorn, 59-62, 91.
Medicinal qualities of the unicorn, 109-110.
Arabian Nights. See A If laila iva-laila.
Aras, 64.
Aristotle, 6, 8, 20, 64, 74, 75, 91, 94, 96, 100, 160.
AsadI, 20.
Asb-i abi. See Faras al-bahr.
Ass:
Indian, 74, 91.
Three-legged, 150.
Wild, 28, 67, 71 pi. ;
20.
'Awfi, 34, 61, 121.
Babur, 11, 13, 17, 46, 89.

(quoted), 81.
Bdbur-ndmah, 46; pi. 32.
Bahram Gur hunting, 37, 39-41, 162; pis. 25, 27.
"Baptistere de St. Louis," 26; pi. 19.

Beatty, A. Chester, mss. in collection, 9, 14, 50, 59, 61, 66, 157; pis. 14, 41, 43.
Berg, Bengt (quoted), 89.
Berlin. Staatliche Museen. Islamische Abteilung:
Indian miniature, 50.
Ivory box, 71.

Pottery fragment, 74, 134, 136; pi. 31.
al-Qazwini ms., 9, 50, 59, 61, 62; pis. 13, 41, 42, 44.
Bestiary, The, 84, 93.
Bezoar stone, 138.
Bible, The, 92, 160.
Bibliotheque Nationale. See Paris. Bibliotheque Nationale.
Bilqis surrounded by the animals of Sulaiman, 47; pi. 34.
Bird friends of the rhinoceros, 23.
al-Biruni, 11, 12, 19, 20, 21, 34, 63, 79, 114, 115, 117, 118, 119, 127, 128,
133, 138-141, 144, 146.

(quoted on color of horns), 115.
(quoted on the khalanj wood), 123.

No. 3 Index
al-BIruni Continued
(quoted on the rhinoceros), 12.
(quoted on the rhino-elephant fight), 90.

Blanford, W. T., 85.
(quoted), 86.
Blood stanched by horn, 61, 127.
Bodleian Library. See Oxford. University. Bodleian Library.
Boston. Museum of Fine Arts:
Manuscripts:
Demotte Shdh-ndmah, 8, 17, 41, 149; pi. 9.
al-Qazwini, 50; pi. 38.

Tile, 27; pi. 19.
Boxes, ivory, 4, 71, 72, 135; pi. 5.
Brazier, 26.
Brehm, Alfred Edmund, 85, 86.
British Museum, ms. in the collection, 6, 8, 9, 24; pi. 7.
Brooklyn. Museum, bedspread, 97; fig. 3.

Brown, W. Norman, 95.
Bucket in possession of R. Stora, 3, 25; pi. 4.
Buckle, silver, 53-55; pi. 39.
Bulan, 67.
Bundahishn, 149-150.
Buzurg ibn Shahriyar, al-Ram-Hurmuzi, 31.
Caesar, Julius, 45, 100.

Cammann, Schuyler, 54, 94, 104, 107, 108, 156.
Candlestick, 27.
Canteen, silver-inlaid, 3; pi. 3.
Carpets:
National Gallery of Art, U.S.A., 28, 160; pi. 21.
Oxford. University. Bodleian Library, 26, 32, 160; pi. 18.
Textile Museum, 160.
Cat with batlike wings, 24.

Cattle, African, 76.
Cervine rhinoceros, 105.
Cervine unicorn, 96, 97.
Chatuq, 116, 123.
Described by al-Kashghari, 122.
Cherif Sabry Pasha. See Sabry, Cherif, Pasha.
Chicago. University. Oriental Institute:
Sculpture from Persepolis, 43, 67, 74, 149; pi. 31.
Ch'i-lin (unicorn), 66, 68, 69, 70, 104, 106, 157; pis. 17, 45.
China's influence on the concept of the unicorn, 102-109.
Chinoiserie, 70, 104, 106, 156.
Cleveland. Museum of Art, ms., 7, 32; pi. 8.
Cochran manuscripts. See New York. Metropolitan Museum of Art.

Colic, cure for, 61.
Convulsions curecf by rhinoceros horn, 133,

Cosmas Indicopleustes, 76, 85, 93.

(quoted), 59, 93.
Couch made of horn and ivory, 104.
Ctesias, 99, 100, 113.
Cups of rhinoceros horn, 111, 112, 131, 132, 133.
Daghestan, sculptured animals of, 17.

al-Damiri, 11, 13, 15, 17, 18, 21, 22, 53, 56, 59, 63, 65, 79, 102, m, 147, 154, 161.
(quoted) 33, 35, 61.
Death of giant animals, 29-31.
Demotte Shdh-ndmah:
Boston. Museum of Fine Arts, 8, 41 ;
pi. 9.
Rockefeller, Mrs. John D., 39, 40, 41.

Design, Mosul, 27, 43, 45, 74; pi. 31.
Designs in horn. See Horns, striae or grooves.
al-Dimashql, 11, 30, 79.
Diodorus Siculus, 76, 77.
Dragon, Chinese. See Lung.
Dyson Perrins collection, 49.
Edelstein, 158.
Elephant:
Attacked by animals, 26-32; pis. 3, 18, 19, 21, 22, 33.
Burial place, 83.
Death of, 30-31; pi. 22.
Enmity of karkadann, 29-31, 78-79, 88, 147; pis. 3, 18, 19, 22, 33.
Enmity to the snake, 47; pi. 35.
Gestation, 52
Legs, jointless, 100-101.
Pregnancy, strange, 16; pi. 16.
Encyclopedic surveys, 152-154.

Erh-ya, 91, 107.
Etymology of the rhinoceros, 7, 75, 94, 106.
Fakhita, 23.
Faras al-bahr, 156; pi. 48.
Faras al-ma', 156; pi. 48.
Feng-huang, 69, 70, 155.
Firdawsi, 18, 19, 37, 154.
Fletcher, Giles (quoted), 125.
Flute sounds of animals, 64-66; pi. 42.
Flying men, 25.

Freer Gallery of Art:
Canteen (41.10), 3; pi. 3.
Manuscripts:
al-Qazwini (07.625), 15, 59, 155; pis. 14, 16.
Shdh-ndmah (30.10A), 39-40; pi. 25.
Mirror, Chinese (29.17, 38.8, 44.5), 105, 108; fig. 4; pi. 48.
Paintings:
Chinese (19.174), 106; pi. 12.

No. 3 Index 197
Freer Gallery of Art Continued

Paintings Continued
Indian (07.625), 10, 15, 59, 155; pis. 14, 16.
(45.29), 50; pi. 37.
(48.8), 48.
Persian (30.10), 39-40; pi. 25.
(38.2), 71; pi. 46.

Turkish (48.17), 69; pi. 45.
Vessel, glass (33-13), 3, J 7, 25; pis. 1, 2.
Friedraann, Herbert, 23.

Friezes, 26-29; pis. 3, 18, 19.
Ganay, Marquis de, ivory box, 4, 24, 136, 137; pi. 5.
Garrett collection. See Princeton University. Library.

Gazelle-horned animal, 95-97.
Geiger, Bernhard, 81, 94, 95, 149.
al-Ghaffari, 127.
al-Gharnati, Abu Hamid, 11, 13, 15, 22, 53, 56, 63, 79, 91, 103, 114.
Giraffe, 15, 155; pi. 16.

Girdle ornaments, 53-55; pi. 39.
Glazer, Sidney, 42.
Goat:
Horns, 61, 62.
Identified with the karkadann, 96.
Goloubew collection. See Boston. Museum of Fine Arts.
Griffins, 3, 4, 24, 28, 67, 152.
Grimms' Fairy Tales, 45.
Guillame le Clerc (quoted), 84.
Gurg. See Wolves.
Gushtasp killing a monster, 36, 40, 162; pis. 24, 26.
"Hai-ma," pi. 48.

Hamd Allah al-Mustawfi al-Qazwini, See al-Mustawfi.
Hare-unicorn, 66 ;
pi. 44.
Harish, 18, 59-62.
Capture of, 60-61; pi. 40.
Description of, 18, 60, 62.

Habitat, 61.
Horn of, 60.
Identification with:
Antelope, 61 ;
pi. 41.
Goat, 61-62; pis. 40, 41.
Snakes, 64; pi. 40.
Wolf, 62; pi. 41.
Karkadann, 96-97.
Medicinal properties of its body, 61.
Virgin-capture of, 61, 92, 95-96.
Harvard College Library, 10, 14, 59, 66.
Heeramaneck, Nasli M., mss. in possession of, 48, 49; pi. 36.
Hemorrhoids, cure for, 131, 139, 142.
14
Heroes slaying the unicorn, 8, 36-37; pis. 9, 23-31.
Herzfeld, Ernst, 44, 67, 74; fig. 1.
Hildegard of Bingen, no.

Hippopotamus. See Faras al-ma'.
Horns:
Antidotal virtue interchange, 136-137.
Arab beliefs concerning, 55, no.
Color of, 62, 115, 139.
Commercial value, 53-58, 117, 139.
Curve in, 21-23.
Description of, 64-65, 99, 114-120, 139.
Detector of poison, 109, 112, 122.
Flute type, 64-66, 150; pi. 42.
Iconography, 3-5, 18-20; pis. 1-20, 22-38, 41-46.

Knot in, 21, 57.
Magical values, 57, 61, no, 133, 160.
Marine variety, 117-120, 127-128, 138-142; pi. 47.
Movable, 76.
Number of, 18-23, 2 7> 61, 62, 64, 67, 70, 79, 91, 108.
Poison, antidotal quality, 99, in, 112, 114, 132.
Poison-detecting quality, 57, 109, in, 112, 114, 122-123, I 3 I > !32, 134.
Portuguese beliefs (quotation), 99.
Regions of export, 176.
Shape of, 18-23, 64, 108, 115, 134-136.
Sharpening, 76, 77, 79.
Size of, 18-23, 98, no, 139.
Strength of, 78.
Striae or grooves, 53, 64, 101, 102, 135, 138, 140, 145.
Use of (by man), 53-58, 67, 77, 98, 102-103, no, 112, 115, 121-127.
Use of (by animal), 67, 92-93.
Host of Sulaiman, 47; pis. 34, 35.
Hsi, 106, 107.
Hsieh-chai, 66, 104, 106, 107, 157; pi. 43.
Hsin T'ang shu (quoted), 54.
Hudud al-'Alam, 11, 55, 114, 116, 118, 119.
Hunting:
Karkadann, 35-46.
Rhinoceros, 45-46; pis.
36, 32, 33.
Unicorn, 36-37, 60, 92-93; pis. 9, 23-31.
Hybrids, 17, 47, 67, 162; pis. 34, 44.
(quotation), 15.
Ibn al-Balkhi, 11, 44, 68, 74, 149.

(quoted), 67.
Ibn al-Faqih, 22.
Ibn al-Suri, 152.

Ibn al-Wardi, 11, 57, no, 132.
(quoted), 131.
:
No. 3 Index 199
Ibn Battuta, 11, 12, 13, 21, 144.

(quoted), 11, 21.
Ibn Bukhtishu', 8, 18, 20, 53, 59, 76, 97, no.
Ibn Fatflan, 11, 13, 35, 45, 53.
Ibn Ibshihi, 15.
Ibn Khallikan, 152.
Ibn Khordadhbeh, n, 14, 15, 20, 48, 53, 55.
al-Idrisi, 15, 53, 152.
Isfandiyar hunting, 36, 39-41, 162; pis. 23, 28-30.
Isidore of Seville, 84.

Iskandar. See Alexander the Great.
Island of Waqwaq, 50; pi. 38.
Ivory, 4, 24, 71, 122-130, 135; pis. 5, 47.
Jahangir, 89, 124; pi. 33.

(quoted), 46.
al-Jahiz, n, 13, 15, 20, 24, 30, 52, 62, 63, 74, 85, 88, 97, 143.
(quoted), 29, 33, 144.
Jamshid, throne of, 31, 67.
Japan, Imperial Treasury (Shosoin), 53-55; pi. 39.
al-Jawhari, 17, 59, 62, 96, 154.
Jazirat al-Tinnin, 66.
Kahler, E., 47; pi. 35.

Kardunn. See Karkadann.
Karg:
Iconography, 37-40; pis. 23-30.
Antelope, 40, 41, 158; pis. 24, 29.
Lion, 17-18, 67; pis. 1, 2, 9, 24.

Wolf, 37-39 5 Pis- 8, 23, 25, 28.
Representations in mss.
Freer Gallery of Art (30.10), 39-40; pi. 25.
Kevorkian, H. (in possession of), 41; pis. 27, 28.
Myers, George Hewitt, 40; pi. 24.
New York. Metropolitan Museum of Art, 41, 158; pi. 29.
Paris. Bibliotheque Nationale, 40; pi. 26.

Princeton University. Library, 41 ;
pi. 30.
Teheran. Gulistan Palace Museum, 39; pi. 23.
Karkadann
Antidotal qualities of its horn, in, 113, 130, 136.
Attacking other animals, 26-34, 7^-79, 88, 147, 160-161 ;
pis. 3, 18-22.
Chinese prototype, 107.

Descriptions by:
al-Biruni, 12.
Ibn FaoMan, 13.
Marvazi, 75.
Enmity to man, 35.

20O Freer Gallery of Art Occasional Papers Vol. i
Karkadann Continued
Features, additional, 52-58, 91.
Fierceness of, 29, 58.
Gestation, 52, 148.

Habitat, 23, 143.
Hoofs of, 64.
Horn of:
Amalgamation with the khutu horn, 151-152.
Antidotal property, 136-138, 161.
Commercial value, 115.
Description, 18-23, 64, 108, 115.
Hunting the, 35-46.
Antelope, 18, 38, 41, 43, 61, 71, 97, 155; pis. 7, 13.
Bovine animals, 13, 14, 97, 149, 159; pis. 10, 11, 13, 15.
Buffalo, 12, 13, 107, 155; pi. 14.
Calf, 23 ;
pi. 15.
Camel, 13.
Donkey, 132.
Elephant, 16, 22; pis. 14, 16.
Giraffelike animal, 47, 155; pi. 34.
Gurg. See Wolf, below.
Harish, 18-19, 62; pi. 41.
Horse, 13, 15, 16, 17, 91, 105, 155, 156; pis. 18-20.
Khutu, 114-120, 136.

Lion, 17-18, 67; pis. 1, 2, 9, 24.
Rhinoceros, 12, 13, 18, 27; pis. 18, 30.
Sharav, 19-20, 31, 34, 62, 97; fig. 3; pi. 13.
Stag. See Antelope.
Wolf, 37-38, 43, 62; pis. 23, 25, 28.
Zebu, 14; pi. 11.

Legs, jointless, 15, 99; pi. 14.
Life span, 52.
Love for the ring dove, 23 ;
pi. 15.
Means of torture in China, 63, 103.
Meat, lawfulness of, 17, 58.
Medicinal values, 110, 131.
Physical appearance, 3-5, 8, 12-27, 47 79> 155-159; fig- 5 J
pis. 1-10,
13-15, 17-38.
Poison-detecting value, 57.
Pregnancy, strange, 15-16, 97, 143, 147, 148.
Reproductions in:
Brazier, 26.
Candlestick, 27.
Carpets, 27, 32, 69, 160; pis. 18, 21.
Manuscripts:
'Ajd'ib al-makhluqdt, 8, 14, 17, 30; pi. 22.
Demotte Shah-namah, 8, 39, 40, 41, 59, 66; pi. 9.
Mandfi'-i hayavdn, 9, 13, 20, 21, 32, 52, 59, 60, 69, 71, 107; pi. 10.

No. 3 Index 20 i
Karkadann Continued
Reproductions in Continued
Manuscripts Continued
Mu'nis al-ahrar, 6-7, 32, 156; pi. 8.
Nizami mss., 41, 47, 48, 49; pis. 34, 36.
al-Qazwini mss.:
Beatty, A. Chester, 14, 157; pi. 14.
Berlin, Staatliche Museen, 9, 19, 31, 59; pi. 13.
Freer Gallery of Art (07.625), 15, 59, 155; pi. 14.
Kevorkian, H. (in possession of), 9, 19, 22, 59; pi. 13.
New York. Metropolitan Museum of Art, 32; pi. 17.
Princeton University. Library, 10, 22; pi. 15.
"Sarre manuscript," 9, 22, 62; pi. 13.

Platter, bronze, 26, 27; pi. 18.
Tile, 26 ;
pis. 18, 19.
Tray, 26, 27.

Vessels, 26, 28; pis. 19, 20.
Reproductiveness, 52.
Scientific and artistic problem, 143-162.
Size, 13, 15.
Tamed, 47"5i; pis. 14, 22, 34, 35.
Thorny tongue, 63, 103, 148.
Voice, 16, 35.
al-Kashghari, Mahmud (quoted), 67, 122.
al-Khazini, 138.
Kevorkian, H. (in possession of) :
Manuscripts:
Nizami, 47; pi. 34.
al-Qazwini, 9, 16, 59, 61, 62, 66; pis. 13, 41, 42.
Shah-namah, 41 pis. 27, 28. ;
Vessel, enameled glass, 28; pi. 20.

Khalanj wood, 102, 123.
Khutu, 1 14-120, 136, 144.
Identified with the rhinoceros, 130-132, 134.
Khutu horn:
Amalgamation of, with horn of the karkadann, 151-152.
Color of, 117-120.
Commercial trade, 121-124.
Poison-detecting quality, 114, 122-123, 131, 138.
Knife handles, 56, 57, 116, 117, 121-127, I 3 I * I 3 2 x 39> >
x 4-6 \ 47-
Knight, William (quoted), 88.
Kolben (quoted), 133.
Konya, sculptured animals of, 4, 26; pi. 3.
Kuhnel, Ernst, 46.

Ku-tu-hsi, 118, 120, 122-123, 130, 133, 134, 136.
Antidotal qualities, 134.
Laufer, Berthold, 9, 54, 59, 96, 100, 101, 103, 107, 116, 118, 122, 130, 133.
(quoted), 102, 103, 112, 122.
Leningrad. Ermitazh:
Shdh-ndmah mss., 42, 43.
Silver-inlaid bronze plate from Kashgar, 26; pi. 18.
Li Hsun (quoted), 145.

Li Shih-chen (quoted), 112.
Lion, 69 ;
pi. 45.
Lion attacking an antelope, 71.

Lion-unicorn, 40, 105; fig. 4; pi. 24.
Lohan, 106 pi. ;

12.
Lorey, Eustache de, plate, 27.

Lung, 69, 70, 105; pi. 45.
Lure of the Shadhahvar's horn, 64-66, 157; pi. 43.
Macy, V. Everit, collection, 161.

Madrasa Muqaddamiya, Aleppo, 4, 35, 51; pi. 6.
Majnun surrounded by animals, 49; pi. 36.

Mandfi'-i hayavdn, 8, 9, 11, 13, 14, 20, 21, 22, 30, 32, 52, 60, 61, 69, 71, 107, 133,
148, 156; pis. 10, 11, 40, 46, 48.

Maqdmdt, 27, 32, 147; pi. 18.
Marvazi, 11, 12, 15, 17, 20, 54, 56, 63, 75, 122, 130, 131, 148.
(quoted), 12, 45, 52, 75.
al-Mas'udi, 11, 15, 53.
Mdtanga-llld, 88.
Meat, lawfulness of, 17, 57-58, 95.

Mihr 'All, ivory carver, 126; pi. 47.
al-Mi'raj, 66-67; pi- 44-
Mirrors, Chinese, 105, 108, 156; figs. 4, 5; pi. 48.

Mohenjo-daro, sculptured animals of, 84, 95.
Monsters, 41, 66, 105; fig. 4; pi. 31.

Mu'nis al-ahrdr ms., 6-7, 32, 156; pi. 8.
al-Mustawfi, 14, 18, 21, 22, 30, 34, 53, 63, 65, 96, 110, 111.
(quoted), 29, 56, 57, 61.
Myers, George Hewitt, ms., 40, 105, 106 ;

pi. 24.
al-Nadim, 11, 53, 55, I44"i45 146.

Narwhal, no, 118, 120, 121, 125, 135, 140; pi. 47.
Na't al-hayaivdn iva-mandfi'uhu, 6, 8, 9, 24; pi. 7.
National Gallery of Art, U.S.A., carpet, 28, 160; pi. 21.
Necklaces, 55.
New York. Metropolitan Museum of Art:
Brazier, 26.
Candlestick, 27.
Luster tile (40.181. 10), 27.
al-Qazwini ms., 24, 32, 68, 104, 158; pi. 17.
Shdh-ndmah ms., 41, 108-109, 158; pi. 29.
Tray (91.1604), 27, 160.
Vessel, bronze, 27.
:
No. 3 Index 203
Nizam! mss.:
Dyson Perrins collection, 49.
Heeramaneck, N. M. (in possession of), 48, 49; pi. 36.
Kevorkian, H. (in possession of), 41, 47; pi. 34.
Sabry, Cherif, Pasha, collection, 49.
Noah and the rhinoceros, 48.
Noldeke, Theodor, 150.
al-Nuwairi, 11, 27, 31, 35, 79, 88, in, 146.
Onager. See Ass, wild.

Oppian, 76.
Ornaments, 53-55; pi. 39.
Oryx, 64, 91, 97.

Ostrich, 155.
Oudh, 89.
Ox, 14, 71, 91 ;
pi. 11.
Oxford. University. Bodleian Library, mss.:

Maqdmdt, 27, 32, 147; pi. 18.
Shdh-ndmah, 50.
Pai-tse, 66.
Pamplona. Cathedral, ivory box in, 4, 24; pi. 5.
Paris. Bibliotheque Nationale, mss.:

Mandfi' al-haywwdn, 8, 59.
Shdh-ndmah, 40; pi. 26.
Paris. Musee du Louvre, 26; pi. 19.
Peacock, 58.
Pelliot, Paul, in.
Persepolis, sculptured animals of, 43, 67-69, 74, 149; pi. 31.
Phoenix. See Feng-huang.
Physiologus, 60, 65, 76, 92, 96, 100, 150.
Pierpont Morgan Library, ms., 9, 14, 59, 60, 61, 71, 93, 107; pis. 10, 11, 40, 48.
Plates, silver-inlaid bronze, 26, 27; pi. 18.
Plato hypnotizes the animals, 48, 65.
Pliny the Elder, 76, 77, 91, 103.
Poison
Antidote against, 99, 109, in, 114, 131, 132.
Means of detection, 109, in, 112, 116, 122, 123, 132, 133, 134.
Polo, Marco, 103.
Pottery:
Garrus, 161.
Mesopotamian, 74, 134, 136; pi. 31.
Persian, 161.
Princeton University. Library, mss.:
al-Qazwini (No. 82G), 10, 14, 23, 62, 64, 66, 104, 135, 158; pis. 15, 40, 42, 43.
Shdh-ndmah (No. 56G), 41; pi. 30.
Tabd'i' al-haya<zudn <iva-kha<wdssuhu, 8.
Pseudo-Callisthenes, 149, 150.

Ptolemy, 75.
: :
Qafd } 6 1.
al-Qazwini, n, 15, 16, 18, 21, 23, 30, 32, 52, 63, 98, 110, 131, 152, 154, 155, 158.
(quoted), 29.
Manuscripts
Beatty, A. Chester, collection, 9, 14, 50, 59, 61, 66, 157; pis. 14, 41, 43.
Berlin. Staatliche Museen, 9, 50, 59, 61, 62; pis. 13, 41, 42, 44.
Boston. Museum of Fine Arts, 50; pi. 38.
Freer Gallery of Art (07.625), 10, 15, 59, 155; pis. 14, 16.
Harvard College Library, 10, 14, 59, 66.

Kevorkian, H. (in possession of), 9, 16, 59, 61, 62, 66; pis. 13, 16, 41, 42.
New York. Metropolitan Museum of Art, 24, 32, 68, 104, 158; pi. 17.
"Sarre manuscript," 9, 14, 24, 62, 63, 64, 65, 66, 135, 156; pis. 13, 16, 41,
42, 44.
Walters Art Gallery, Baltimore, 8, 10, 14, 16, 17, 48, 65, 156, 161; pi. 22.
Quinsy, cure for, 61.
Rhino-birds, 23.
Rhinoceros
Antagonism to the elephant, 78-82.
Attacking man, 35-36.
Calf, care of, 63.
Commercial use of, 53-57.
"Communicating with the sky," 145.
Description by:
al-Birunl, 12.
Classical World, 74-78.
Cosmas Indicopleustes, 59.
Li Hsiin, 145.
Timothy of Gaza, 75.
Domesticated, 48.
Elephant-rhinoceros fight, 27, 30, 78-90; fig. 2; pis. 18, 33.
Fierceness of, 147.
Habitat, 87.
Horns, 72, 78, 91, 160.
Hunting the animal, 36, 45-46; pis. 32, 33.
Hypnotized, 48.
Bovine animals, 48, 106, 108; pi. 12.
Horse, 17, 91; pis. 18-20.

Khutu, 130.
Stag, 18, 97.
Legs, jointless, 15, 100.
Magical remedies, 57.
Meat, lawfulness of, 57-58, 95,
Prickly tongue, 103.
Representations in art:
Carpets, 27, 146; pi. 18.
Fresco, 78.
Mirrors, 108; fig. 5.
:
No. 3 Index 205
Rhinoceros Continued
Representations in art Continued
Paintings, 157; pis. 12, 32, 33.
Statuette, 151.
Species:
African, 35, 79, 80, 146.
Indian, 86, 87, 146.
Siberian, 20.
Tamed, 48, 49; pis. 14, 36.
Voice, 35.
Zoological opinion, 85-86.
Rhinoceros horn:
Magical powers, 99, 132, 133, 160.
Sharpening of, 77.
Size, 98.
Strength, 78.
Striae, grooves, etc., 53, 102, 135, 145.
Use (by man)
Cups, in-112, 131, 132, 133.
Ornaments, 53-55, 146; pi. 39.
Sword hilts, 115.
Throne decoration, 103.
Valuation, 53-58.
Ring dove, 23 pi. 15.
;
Rockefeller, Mrs. John D., ms., 39, 40, 41.

Rosenthal, Franz, 60, 65, 75, 114.
Rsyasrnga, legend of, 95-96.
Rukh, 19, 33-34, 83.
Sabry, Cherif, Pasha, ms., 49, 151-152.

Sarre, Friedrich, 4, 26, 43, 44, 67; pi. 3.
"Sarre manuscript," 9, 14, 24, 62, 63, 64, 65, 66, 135, 156; pis. 13, 16, 41, 42, 44.
Sauvaget, Jean, 5.
Sculpture, stone, 43, 67-68, 74, 109; pis. 3, 31.
Sea-horse, 156; pi. 48.
Sea-Serpent Island. See Jazirat al-Tinnin.

Shadhahvar, 64, 65, 104, 157; pis. 42, 43.
Shdh-ndmah mss.:
Boston. Museum of Fine Arts, 8, 17, 41, 149; pi. 9.
Freer Gallery of Art (30.10A), 39-40; pi. 25.
Kahler, E., 47; pi. 35.
Kevorkian, H. (in possession of), 41; pis. 27, 28.
Leningrad. Ermitazh, 42, 43.
Myers, George Hewitt, 40; pi. 24.
New York. Metropolitan Museum of Art, 41, 108-109, I 5%> pi- 2 9-
Oxford. University. Bodleian Library, 50.

Paris. Bibliotheque Nationale, 40; pi. 26.
Princeton University. Library (No. 56G), 41, 158, 162; pi. 30.

2o6 Freer Gallery of Art Occasional Papers Vol. i
Shdh-ndmah mss. Continued

Rockefeller, Mrs. John D., 39, 40, 41.
Teheran. Gulistan Palace Museum, 39; pi. 23.
Sharav, 19, 62, 97; fig. 3; pi. 13.

Attacking other animals, 31, 34.
Shipley, Sir Arthur (quoted), 76.
Shuba, 21.
Silkworm, 57.
Simurgh, 8, 32, 56, 69; pis. 8, 17, 18, 34-37.
Sinad:
Horn, 64.
Identification with bovine animals, 63 ;
pi. 42.
Identification with the karkadann, 62-64.
Pregnancy, strange, 16, 62; pi. 16.
Tongue, thorny, 63.
Sindbad, voyages of, 30, 31, 83.

SIranis, 65.
Snakes, 62 ;
pi. 40.
Sohn-Rethel collection, 46, 89; pi. 33.
Solinus, 77, 91.

Spenser's Faerie Queene (quoted), 45.
Sphinxes, 3, 24, 28.
Stone, George Cameron, 127.
Stora, R., 3, 25 ;
pi. 4.
Stott, Ken (quoted), 80, 87.

Strabo, 100.
Sulaiman and the animals, 47; pis. 34, 35.
Symbolism of the animals, 157.
Talking tree, 50-51, 151; pis. 6, 38.

Talmud, The, 59.
Tang hui yao (quoted), 54.
Ta'rikh-i Mubdrak-Shahi (quoted), 45.
al-Ta<ivhidi, Abu Haiydn, 11, 18, 92, 96.
(quoted), 60.
Teheran. Gulistan Palace Museum, ms., 39; pi. 23.
Tell Asmar seal, 83, 84; fig. 2.

Textile Museum. Washington, D. C, 160.
Thrones of horn and ivory, 104.
Thunberg, Karl Peter (quoted), 132.
Tiles, 26, 27; pis. 18, 19.
Timothy of Gaza, 75, 84.
Togan, Zeki Validi, 60, 118.
Torture in China, 63, 103.
Trade routes, 128-129.
Tramontana, 80, 81.
Trays, 26, 160.
Turks, influence of, 121-124, 134.
:
No. 3 Index 207
'Ubaid Allah ibn Jibril ibn Bukhtishu', 6, 11.

Unicorn:
"Apocryphal," 67-70; pis. 17, 45.
Description, 3-5, 6, 8, 10, 17, 70, 109-142; pis. 1-6.
Horn, 120; pi. 47.

Hunting the animal, 36-37, 60, 92, 93.
Iconographic problem, 3-5; pis. 1-6.
Identification with a kid, 92.

Literature cited on, 163-191.
Lore, influenced by, or found in:
Ancient Orient, 73-74; pi. 31.
Bible and Christian literature, 92-93.
China, 68, 101-109.
Classical world, 74-91.
India, 81, 94-101 ;
fig. 2.
Medieval world, 109-142.

Muslim civilization, 1-72, 91, 137-138.
Renaissance, 84.
Seljuk period, 74.
Medicinal qualities, 57, 61, 109-111, 113, 130, 133.
Unicorn-elephant fight, 3, 28, 84; pis. 3, 18, 19, 21, 22.
Varieties of:
Classical
Horse, 91.
Indian ass, 74, 91.
Oryx, 64, 91, 97.
Rhinoceros, 74-93.
Far Eastern:
Ch'i-lin, 66, 68, 69, 70, 104, 106, 157; pis. 17, 45.
Hsi, 107.
Hsieh-chai, 66, 104, 106, 107, 157; pi. 43.
Ku-tu-hsi, 118, 120, 122-123, 130, 133, 134, 136.

Pai-tse, 66.
Marine, 120-121.
Muslim:
Antelope, identified with, 18, 38, 41, 43, 61, 71, 97, 155; pis. 7,
13, 26, 27, 29.

Buffalo, identified with, 12, 13, 107, 155; pi. 14.
Bulan, 67.
Camel, identified with, 13.
Chatuq, 116, 122.

Harish, 18-19, 60-62; pi. 41.
Horse, identified with, 13, 15, 16, 17, 155, 156; pis. 18-20.
Karg, 37-41,; pis. 23-30.
Karkadann, 13-17, 47-72.
Khutu, 114-120, 136, 144.
Lion, identified with, 17-18, 62, 67; pis. 1, 2, 9, 24.
al-Mi'raj, 66; pi. 44.

Shadhahvar, 64, 65, 104, 157; pis. 42, 43.
: :
Unicorn Continued
Varieties of Continued
Muslim Continued
Sharav, 19-20, 31, 34, 62, 97; fig. 3; pi. 13.
Sinad, 62-64.
Wild ass, 28, 67, 71; pi. 20.
Wolf, identified with, 37-41, 43, 62; pis. 23, 25, 41.
Virgin-capture of, 60, 92.
"Water-conning," 150.
University of Chicago. See Chicago. University.
Valle, Pietro della, 121.

Venice, Treasury of San Marco, 120; pi. 47.
Vessels
Bronze, 3, 25, 26, 27; pis. 4, 18, 19.
Glass, 3, 17, 25, 28; pis. 1, 20.
Pottery, 74, 134, 136, 161; pi. 31.
Silver-inlaid canteen, 3 ;
pi. 3.
Virgin and the unicorn, 60-61.
Walrus, 125, 129.

Walrus tusk, 120, 121, 124, 135; pi. 47.
Walters Art Gallery, Baltimore:
Ivory knife handle, 56; pi. 47.
Manuscripts
'Ajd'ib al-makhluqdt, 14, 17, 48, 65, 156, 161; pi. 22.
Bdbur-ndmah, 46; pi. 32.
al-Qazwini, 8, 10, 14, 16, 17, 48, 65, 156, 161; pi. 22.
Wang Pi-chih, 102.

Waqwaq island, 50, 55.
"Water conning," 150.
Water horse. See Faras al-ma'.
Widener collection. See National Gallery of Art, U.S.A.
Williamson, Capt. Thomas, 82.
Winged cat, 24.
Winged equine beasts, 44; fig. 1.
Winged griffon, 70.
Winged karkadann, 24-25, 26-27, 3 2 > 47> I 5^> I S9\ pl s - I_ 5> 7"9> I _
7 2i, 25, 31, 35.
Winged lion, 3, 25 pis. 1, 2. ;
Winged men, 25.

Winged rabbit, 25.
Winged unicorn, 3, 4, 8, 28, 43; pis. 1-5, 7-9, 17-20, 25, 31, 35.
Wisu, 128.
Wolves, 37-41, 43, 62; pis. 23, 25, 28, 41.
Yahmur, 71 ;
pi. 46.
Yak, 103.
No. 3 Index
Yale (mythical animal), 76.

Yaqut, 152.
Zafar-ndmah, 45-46.
al-Zamakhshari, 17, 58, 147.
(quoted), 58.
Zebu, 14; pi. 11.
1
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. I. NO. 3. PLATE 2
Unicorns on glass vessel of plate 1.

FREER GALLERY OF ART OCCASIONAL PAPERS VOL. I. NO. 3, PLATE 3
Upper: Details of a bronze canteen. Freer Gallery of Art, No. 41.10.

Middle and lower: Sculptures from Konya (after 225, fig. 12, and 227, pi. 17).
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. 1. NO. 3. PLATE 6
Plaque, Madrasa Muqaddamiya, Aleppo. Photograph courtesy of M. Jean

Sauvaget.
VOL. NO. 3. PLATE 7
FREER GALLERY OF ART OCCASIONAL PAPERS
I.
"Kardunn" from a Na't al-hayawdn manuscript. Courtesy of the Trustees of

the British Museum.
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. I, NO. 3. PLATE 10
"Karkadan" from a Mandft-i hayavan manuscript. Photograph courtesy of The

Pierpont Morgan Library.
"Domestic Oxen" from a Mandfi'-i hayavdn manuscript. Photograph courtesy of

The Pierpont Morgan Library.
Lohan on a rhinoceros. Freer Gallery of Art, No. 19.174.

FREER GALLERY OF ART OCCASIONAL PAPERS VOL. !, NO. 3, PLATE 13
Representations of the karkadann. Upper left: Photograph courtesy of

Islamische Abteilung, Staatliche Museen, Berlin. Upper right: Courtesy of
H. Kevorkian, New York. Lower: Photograph courtesy of the owner.
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. I, NO. 3, PLATE 14
Representations of the karkadann. Upper: Courtesy of A. Chester Beatty, Esq.,

London. Lower: Freer Gallery of Art, No. 07.625. foL<3 H6S"v/erô
Upper: "Shutur-gav-palang." Freer Gallery of Art, No. 07.625. Lower: "Sinad."

Courtesy of the owner.
Right title page of an al-Qazwini manuscript. Courtesy of the Metropolitan

Museum of Art.
The karkadann pursuing an elephant. Upper: On a tile. Photograph courtesy

of Islamische Abteilung, Berlin. Middle: On a bronze plate. Hermitage, Lenin-
grad. Lower: On a carpet. Courtesy of the Bodleian Library, Oxford.
Enameled glass vessel. Photograph courtesy of H. Kevorkian, New York.

Mughal carpet. Photograph courtesy of the National Gallery of Art, Washington

(Widener collection).
The death of the elephant. Photograph courtesy of the Walters Art Gallery,
Baltimore.
Isfandiyar killing two karg monsters. Teheran, Gulistan Palace.

Upper: Gushtasp killing a karg. Lower: Iskandar killing a karg. Photographs

courtesy of George Hewitt Myers, Washington, D. C.
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. 1, NO. 3. PLATE 26
Gushtasp killing a karg. Bibliotheque Nationale, Paris.

FREER GALLERY OF ART OCCASIONAL PAPERS VOL I, NO. 3. PLATE 28
Isfandiyar killing karg monsters. Courtesy of H. Kevorkian, New York.

FREER GALLERY OF ART OCCASIONAL PAPERS VOL. !. NO. 3. PLATE 29
Isfandiyar killing karg monsters. Courtesy of the Metropolitan Museum of Art.

Isfandiyar killing karg monsters. Courtesy of Princeton University Library.

Babur on a rhinoceros hunt. Photograph courtesy of the Walters Art Gallery.

Jahangir on a rhinoceros hunt. O. Sohn-Rethel collection, Diisseldorf (after 156,

p. 418).
Bilqis surrounded by animals. Courtesy of H. Kevorkian, New York.

FREER GALLERY OF ART OCCASIONAL PAPERS VOL. I. NO. 3, PLATE 35
Sulaiman's host. Courtesy of Dr. E. Kahler, Princeton.

Laila visiting Majnun. Photograph courtesy of N. Heerameneck, New York.

The animal kingdom. Freer Gallery of Art, No. 45.29.

The Island of Waqwaq. Courtesy of the Museum of Fine Arts,

Boston.
FREER GALLERY OF ART OCCASIONAL PAPERS VOL. !. NO. 3, PLATE 40
Upper: Snake harish. Courtesy of Princeton University Library. Lower:

The capture of the harish. Photograph courtesy of The Pierpont Morgan Library.
The harish in al-Qazwini manuscripts. Upper: Courtesy of the owner.

Middle left: Photograph courtesy of Islamische Abteilung, Berlin. Middle
right: Courtesy of H. Kevorkian, New York. Lower: Courtesy of A. Chester
Beatty, Esq., London.
Upper: "Shadhahvar." Photograph courtesy of the owner. Middle: "Sinad."

Courtesy of Princeton University Library. Lower left: "Shadhahvar." Courtesy
of H. Kevorkian, New York. Lower right: "Shadhahvar." Photograph courtesy
Islamische Abteilung, Berlin.
The lure of the shadhahvar's horn. Upper: Courtesy of Princeton

University Library. Lower: Courtesy of A. Chester Beatty, Esq.,
London.
The Mi'raj in al-Qazwini manuscripts. Upper: Photograph courtesy of
Islamische Abteilung, Berlin. Lower: Courtesy of the owner.
Yahmur," from a Mnndfi'-i hayavan manuscript. Freer Gallery of Art, No.

38.2.
Upper: Detail of lower "Unicorn horn" shown below. Middle: "Unicorn

horns." Treasury, San Marco, Venice (after 238, figs. 211 and 212). Lower:
Knife with handle of walrus ivory. Photograph courtesy of the Walters Art
Gallery.
Upper: "Hai-ma" on bronze mirror. Freer Gallery of Art, No. 44.5.

Lower: "Asb-i abi" from a Manaft -i hayavan manuscript. Photograph
courtesy of The Pierpont Morgan Library.
LIBRARIES
SMITHSONIAN INSTITUTION
3 9088 01575 4989

The Unicorn in Muslim Traditions

Uploaded by

Copyright:

Available Formats

The Unicorn in Muslim Traditions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Unicorn in Muslim Traditions

Uploaded by

Copyright:

Available Formats

What is the publication about?

What is the publication about?

What illustrations are included in the publication?

What illustrations are included in the publication?

= SMITHSONI AN INSTITUTIONS

FREER GALLERY OF ART

Studies in Muslim Iconography

Articles dealing with objects in the Freer Collection and involving

FREER GALLERY OF ART

Studies in Muslim Iconography

scripts in the India Office Library and the British Museum.

in thisvolume, namely, one plate which is specifically identi-

1 929-1940) were vested in 1949, pursuant to law. The use

The transliteration of Arabic used in this investigation

List of plates vii

List of text illustrations xii

The iconographic problem 3

1. Glass vessel, gilded and enameled. Probably made for a Rasulid

of the thirteenth century. Freer Gallery of Art, No. 41.10.

9. "Iskandar Battling a Karkadan" from the "Demotte Shah-ndmah"

10. "Karkadan" from a Manafi-i hayavdn manuscript. Iran (Ma-

12. Lohan on a rhinoceros, detail of the painting "Lohans Moving

the left miniature.

of the thirteenth century. Leningrad, Hermitage Museum.

1300. Paris, Musee du Louvre.

No. 3 List of Plates ix

Lower: Detail of a luster tile showing karkadann pursuing an ele-

phant. Iran, thirteenth century. Boston, Museum of Fine Arts.

26. Gushtasp killing a karg, from a Shah-ndmah manuscript. Iran,

1497 (902). New York, possession of H. Kevorkian.

30. Isfandiyar killing karg monsters, from a Shah-ndmah manuscript.

32. Babur on a rhinoceros hunt. Mughal India, second half of the

34. Bilqis surrounded by the animals of Sulaiman, from a Nizami

From Shdh-namah manuscript, now dispersed. Iran, 1497- 1504

persed. Iran, sixteenth century. Boston, Museum of Fine Arts

42. Upper: "Shadhahvar," from the "Sarre al-Qazwini manuscript."

Middle: "Sinad," from an al-Qazwini manuscript. Iran, seven-

44. The mi'raj in al-Qazwini manuscripts.

for a more extensive investigation. As a result not only has

remains a certain amount of mystery. This applies especially

the animals portrayed in a heraldic style and posture (pi. 4).

Here head and body are basically equine, although many

horn had he wanted to do so. At this stage of the investiga-

THE IDENTIFICATION OF THE ANIMAL

Na ( t al-hayawan wa-mandfi'iihii in the British Museum, prob-

See also 62, pp. 34-35, and figs. 34-36.

(99, p. 109, footnote; Ibrahim b. Wasif-Shah, in 98, vol. 1, p. 160, footnote 1;

caption; and al-Idrisi in 98, vol. 1, p. 180; jl J? kazkazan, in the Cairo

this time with a feline body is to be found on a page of the

6, fol. 134b; 7, fol. 88b; 201, vol.

jy khartut occurs in al-Akfani (277, p. 354) ;

In a ribbonlike composition four animals are shown, one in

karkadan. Our third example is in a miniature of the "De-

trated scientific manuscripts such as al-Qazwini's 'Ajd'ib al-

makhluqdt or the various works connected with Ibn Bukh-

54, p. 106, No. 33.

jl j*^A Na't al-hayaivdn, which is attributed to Aristotle, while the dis-

hayawdn of Ibn Bukhtishu'. Both were put together by an anonymous com-

No. 3 I. The Unicorn Ettinghausen 9

examples of this animal, since these manuscripts have been