tt

we
i
 Digitized by the Internet Archive
in 2022 with funding from
Kahle/Austin Foundation

https://archive.org/details/historyofboardga0000murr_p4b9
 A HISTORY OF
BOARD-GAMES
OTHER THAN CHESS
 eta tS ORY GO

BOARD-GAMES
OTHER THAN CHESS
BY

H. J. R. MURRAY

HACKER ART BOOKS, INC.

NEW YORK
RODRS
 First published 1952 by
Oxford University at the Clarendon Press
Reissued 1978 by
Hacker Art Books, New York.
Library of Congress Catalogue Card Number 77-73728
ISBN 0-87817-211-4
This reprint has been authorized by the Oxford University Press.
Printed in the United States of America.
 PRE Ab
Tuts work completes the survey of board-games which I began in my
History of Chess (Oxford 1913). The aim of this survey has been to
present as complete a record as is possible of the board-games that exist,
or have existed, in different parts of the world, to establish how they
are played, to sketch what is known of their histories and developments,
and to investigate the ultimate origin of board-games in general and the
circumstances which made possible their invention. The only games
that I have deliberately omitted are those of recent invention in Europe,
America, and Japan which exhibit no new features of aim or rule.
These are nearly always games of chance of the class I term ‘race-games;
their number is great and their lives are usually short.
Many monographs on the board-games of a single country or people
have been published in the last fifty years, but this is the first attempt to
make a comparative study of the board-games of the whole world. It
is naturally based on these monographs and on other descriptions of
~ games in the journals of learned societies or briefer mentions of games
in books of travel. I give later a list of the books and articles which I
have consulted, and acknowledge my indebtedness to their authors. I
have been careful to give in my account of each game the source of my
information. My own contribution has been the adoption of a uniform
system of terms in describing the games.
While it is possible to deal with many European games, such as
merels, draughts, hnefatafl, rithmomachy, fox and geese, and tables, on
the lines which I had adopted for chess, it soon became evident that the
majority of games could only be described as they are played today.
This explains a certain lack of proportion in the treatment of different
games.
I should never have written this book without the help which I have
freely received from many persons who have been interested in board-
games. First among these was Dr. A. C. Haddon who, while I was still
occupied in writing my previous book, urged me to write next a book
on similar lines about other board-games, put at my service the material
which he had collected about them, and continued td add to it so long
as we were neighbours in Cambridge. I have also been fortunate in
receiving much new and unpublished material from others during the
past twenty years. Mr. N. Teulon Porter gave me all the valuable
vi PREFACE
material which he and other students under Dr. Haddon in the Cam-
bridge School of Ethnology had collected for a projected thesis on the
game of mancala. Among those who had helped in making this collec-
tion were Mr. R. de Z. Hall and his brother (Tanganyika), Dr. L. S. B.
Leakey (Uganda), and Mr. H. A. Stayt (Transvaal). I refer to this
collection as CP. (Cambridge Papers). Prof. H. T. H. Piaggio sent me
information which he had obtained from his pupils at Nottingham,
Mr. Hilmi Samara (Arab games in Palestine), and Mr. M. S. Kadri
(Egypt). My son, Mr. K. C. Murray, Surveyor of Antiquities in
Nigeria, has given valuable help about Nigerian games, including a
number of scores of games which he had taken down from actual play,
and a collection of Benin games which had been observed by Mr. J. U.
Egharevba, curator of the Benin Museum. Mr. A. G. Shirreff, I.C.S.,
has also put at my service his collection of games played in northern
India, which has been of the greatest value to me in filling a serious gap
in my information.
Among those who have given me important help are Prof. R. G.
Austin (for classical games of Greece and Rome) who also has helped
me with the proofs and with the general arrangement of this book,
Mr. H. T. Braunholtz, C.B.E., of the British Museum (for mancala
games), Mrs. Danielli (for Malagasy games), Mr. H. D. Dickinson
(for general advice), Mr. C.J. Gadd of the British Museum (for Assyrian
and Sumerian games), Miss B. J. Kirkpatrick, Librarian of the Royal
Anthropological Institute (for bibliography), Dr. F. R. Lewis (for
tawlbwrdd), Prof. R. A. S. Macalister (for ancient Palestinian games),
my daughter Miss K. M. E. Murray (for boards in cloisters and on
tombs), Mr. H. O. Robinson and Mr. W. W. Skeat (for Malay games),
Miss L. S. Sutherland (for pentalpha), and Prof. Tsuboi (for Japanese
games). To all of these and others I express my most grateful thanks
for their help.
In conclusion, once again I should like to express my personal
gratification that this book is appearing from the same University Press
which published the pioneer work on board-games, Thomas Hyde’s’
De Ludis orientalibus in 1694.
H.J.R.M.
HEYSHOTT
I9SI
 CONTENTS
LIST OF ILLUSTRATIONS page ix

ABBREVIATIONS page xii

CHAPTER 1. INTRODUCTORY page 1

General characteristics of board-games. Sources of information. Classes of board-
games. Boards, lined and latticed. Pieces or men. Lots and dice. Powers of move
used in board-games. Methods of capture.

CHAPTER 2. GAMES IN THE ANCIENT WORLD page 12

The el-Mahasna board. 1. Egypt; 2. Ur; 3. Assyria; 4. Palestine; 5. Cyprus; 6. Crete;
7. Greece; 8. Rome; 9. Celtic games; ro. India; 11. China.

CHAPTER 3. GAMES OF ALINEMENT AND CONFIGURATION

Merels. Shiva. Go-bang. Halma. Salta. page 37

CHAPTER 4. WAR-GAMES page $3

Battle-games; petteia; tafl. American Indian games. Alquerque; draughts; atlanbaj;
chess; rithmomachy; fanorona. Territorial games; wei-k’i. Blockade games. Clear-
ance games. War-games played with dice; tab.

CHAPTER 5. HUNT-GAMES page 98

Games on the alquerque boards. Chinese andJapanese soldier games. Fox and geese,
ancient and modern. Leopard games. Tiger games.

CHAPTER 6. RACE-GAMES page 113

Alea and its affiliated games in Asia. Tables (backgammon, trictrac, puff and verkeer,
&c.). Race-games on square boards; on cruciform boards. Single-track games.
American Indian games. The Arabic astronomical game.

CHAPTER 7. MANCALA. PART I page 158

Distribution and age of mancala. General characteristics. Mancala II games in Asia,
Africa, and America.

CHAPTER 8. MANCALA. PART UII page 205

Mancala III games. Mancala IV games in Africa.

CHAPTER 9. THE DISTRIBUTION AND ORIGIN OF

BOARD-GAMES page 226
Lack of detailed information. Vitality of games. Their spread in the Old World
explained. Independence of the American Indian games. Previous views on the
origin of games examined. Necessary conditions for their invention. Most games
the result of experimentation with existing objects. The analogy of string-games.
Vili CONTENTS
TABLE SHOWING DISTRIBUTION OF BOARD-GAMES page 240

BIBLIOGRAPHY page 242

INDEX OF GAMES page 249

INDEX OF GAMES BY COUNTRIES page 258

GENERAL INDEX page 261

 BIST OPCIELUS
TR ero Ns
Gaming-board from Ur p. xiv
Fic. 1. Pre-dynastic board from El-Mahasna 13
2. From tombs at Benihassan 14
3. Egyptian race-game. XI and XII Dynasties 15
4. Egyptian game-box. Upper face 16
5. Egyptian game-box. Lower face 17
6. Board from Ak-hor, Cairo Museum 17
7. Game diagrams on the roofing slabs of the
temple at Kurna 19
8. Game-board from the royal graves, Ur 20
9. Board from Ur (B.M. 123331) 21
10. Terra-cotta plaque from Gezer 23
11. Game-board from Enkomi (Cyprus) 28
12. Board from Knossos (Crete) 24
13. From a terra-cotta, Athens 26
14. Pentalpha, Crete 28
15. Board for Duodecim Scripta (from Timgad) 30
16. Board for Duodecim Scripta (from Ostia) 31
17. Zeno (White) to play 2, 5, 6 32
18. Boards for three-in-a-row games 38
19. Halma 51
20. Game-pieces (after Montelius) 58
21. Wimose board 58
22. Board from Gokstad ship 58
23. The Ballinderry game-board 59
24. Hnefatafl pieces 60
25. Alea Evangelii 62
26. Tablut, Lapland 63
27. Alquerque 65
28. Mogol Putt’han 67
29. Perali kotuma 68
30. Terhtichii 68
31. Quadruple Alquerque 69
314 Boards for Indian war-games 70
31b Pretwa 70
32. Awithlaknannai 71
33. Dama (Philippines) 79
 LIST OF ILLUSTRATIONS
. Board from Liberia
. Rithmomachy
. Fanorona
. Wei-k’i. Various captures
. Wei-k’i. The White group is safe from attack
. Wei-k’i. Ko
. Ko-no
. Mu-torere
. Chuki
. The Hare game
. Solitario (Peru)
. Sixteen soldiers (China and Japan)
. Yeung luk sz’ kon tseung kwan
. Fox and geese
. Solitaire board
. Fox and geese (Venice)
. Hat diviyan keliya
. Pulijudam
. Rimau-rimau
. Board and notation for tables
. Ashtapada
. Saturankam
. Ashta-kashte
. Sadurangam
. Gavalata
. Chaupur; Pachisi
. Dajam
. Asi keliya
. Ludo
. Len doat (Siam)
. Single-track boards (after Marin)
. Kawada keliya
. Nyout
. Chinese promotion game
. Patolli
. Board from the Fejervary Codex
. Stone board from Oraibi
. Stone board from Oraibi
. Stone boards for Totolospi
. Zohn Ahl
 LS TORS Si ReATONS xi
Fic. 74. Quince 156
75: Astronomical game 157
76. Mancala board (Hyde) I61
77- Pallankuli board (Tamils, S. India) I6I
78. Wari board, West Africa (Anthropological
Museum Marischal College, Aberdeen) 162
79. Ayo board (B.M. 1928, 2-7, 1) 162
80. Ceylon boards I7I
81. Bronze board from Benin 177
82. Aghadaghada. Benin 181
83. Urise. Benin 190
84. Board for Gabata (Abyssinia) 206
85. Ochi board for mancala IV 207
86. Mefuhva. BaVenda tribe, N. Transvaal 211
 ABBREVIATIONS
a. ante (in dates).
Alf. Alfonso MS. (Das Spanische Schachzabelbuch des K. Alfons des
Weises. Leipzig 1913).
BAM. Bulletin de ?Académie Malgache.
BS. Bonus Socius MS. (see HC., 619-42).
c. circa (in dates).
GB: Civis Bononiae MS. (see HC., 643-99).
CP. Cambridge Papers (see Preface).
CPB. Corpus Poeticum Boreale, ed. York-Powell, Oxford 1891.
Cit. Citolini, La Tipocosmia, Venetia 1561.
Cz Czech.
Dan. Danish.
Du. Dutch.
EETS. Early English Text Society.
ENI. Encyclopedie van Nederlandsch Indié, iv, 1921.
Encyc. Méth. Encyclopédie Méthodique. Dictionnaire de Jeux, 1792.
Ger. German.
HC. H. J. R. Murray, A History of Chess, Oxford 1913.
Hung. Hungarian.
TAFE. Internationalische Archiv fiir Ethnographie.
ic Icelandic.
ILN. Illustrated London News.
Isl. Gatur. — fslenzkar Gatur, Kaupmannahéfn 1888-92.
It. Italian.
JAI. Journal of the Anthropological Institute.
JASB. Journal of the Asiatic Society of Bengal.
JHS. Journal of Hellenic Studies.
JRAI. Journal of the Royal Anthropological Institute.
JRAS. Journal of the Royal Asiatic Society.
K. MS. British Museum, Kings 13 A.xviii.
Lat Latin.
MDa. Middle Danish.
MDGNVO. Mittheilungen der deutschen Gesellschaft fiir Natur- und Volkerkunde Ost-
asiens.
MDu. Middle Dutch.
ME. Middle English.
MF. Middle French.
MHG. Middle High German.
ML. Medieval Latin.
MSw. Middle Swedish.
 ABBREVIATIONS
Museum ftir Volkerkunde, Berlin.
North American Indians.
Old English.
Oxford English Dictionary.
Old High German.
Palestine Exploration Fund. Quarterly Statement.
Plate.
Polish.
Provengal.
Royal Historical Society.
Russian.
Sudan Notes and Records.
Swedish.
Zeitschrift der Deutschen Morgenlandischen Gesellschaft.
 |
INTRODUCTORY
Gags, which resemble chess, draughts, and backgammon in being played on
a specially arranged surface with pieces or ‘men’, whose powers of move and
capture are defined by the rules of each game, are designated as “board-
games’, German Brettspiele. The surface of play is commonly known as
the ‘board’, although it is not necessarily made of wood or other material,
and is, in fact, in many parts of the world more often marked out on a level
piece of ground whenever a game is about to be played. The aim of this book
is to collect all the available information concerning the existence of board-
games other than chess! in any part of the world, to classify the games by the
activities which they typify, to describe how each game is played, and to set
out what is known about their histories.
Most board-games are played by two persons or by two opposing sides of
two or more persons, but some are played by three, four, or more persons
playing individually, and a few are played by a single person. Each player
has a given number of men, which in the great majority of games are un-
differentiated except to distinguish the opposing forces. In some games the
men are arranged on the board before play begins and the initial arrangement
is fixed; in others the players are free to adopt their own arrangement and
begin by entering their men, sometimes in alternate turns of play, sometimes
simultaneously; in others again the players, after entering some of their men,
may retain the rest ‘in hand’ and enter them later in a turn of play. In most
games, after the opening play is ended, the game proceeds by alternate turns
of play by the opposing sides, and the first player to achieve the aim of the
game wins.
Men have played board-games for more than 4,000 years, but for the
first half of this period we are dependent on a few drawings of games in
progress (such as may be seen on the walls of ancient Egyptian temples and
tomb-chambers), a few references or allusions to games in ancient literature,
and the game-materials which have survived to modern times. None of these
throws much light on the nature or method of play of these ancient games.
Drawings of games establish that their artists knew that games were played
by their contemporaries, but nothing more, and medieval and even modern
pictures of games of which we have complete knowledge warn us that it is
unsafe to rely on them for accuracy of detail.2 The information given by
T Already treated in my History of Chess, Oxford 1913.
2 Thus, a French game-manual of 1686 shows two persons playing chess on a board of 3 x8
cells, and Matisse, in his Portrait de Famille, shows two children playing Polish draughts on a board
of 117 cells.
5356 B
2 INTRODUCTORY
drawings is also limited by the artistic development of their time. Thus, the
ancient Egyptians drew in profile only, and their drawings tell nothing about
the width of the board or the transverse arrangement of the pieces.
Allusions to games were made for contemporaries and assume a know-
ledge of the games mentioned, without which they convey very little to any
reader who is ignorant of the games.
Game-materials consist of boards, pieces, and dice or other implements of
chance. None of these by itself necessarily implies that they were used for a
game, though this is more likely when they are found together. A board,
even if it shows the same pattern of lines or cells as a known game-board,
may have served other purposes before it was associated with a game, and
the same is true of the simpler forms of pieces. Implements of chance by
themselves establish nothing, since they have been used from the earliest
times for divination or simple gambling. Moreover, these remains are scanty,
difficult to date, and often inconclusive. So long as a people is content to
mark out the board on the ground and to use any convenient object at hand
for the pieces, there is very little chance of their survival. Even in lands where
special boards and men are used, very few complete game-sets survive for
more than 400 years: it is a curious irony of fate that most of the European
chessmen and tablemen that have survived to our own day are those which
were lost in the Middle Ages. It is only in lands where it was customary to
bury a man’s possessions with him for use in his life after death, that we have
satisfactory and datable evidence for the early existence of board-games. In
this way we learn that board-games were played in ancient Egypt, Mesopo-
tamia, Assyria, Cyprus, and Crete, the evidence for which will be summa-
rized in Chapter 2.
It was not until the early centuries A.D. that scholars showed much interest
in the nature of games, and Byzantine antiquaries began to try to elucidate
the references to games in the older Greek literature. Unfortunately they
assumed that these games were dead and knew nothing about the games of
other peoples. They were, accordingly, in no better position to explain the
allusions than scholars of our own day, and have only succeeded in adding
to the obscurity of the allusions. This will be made clear in the following
chapter.
It was not until the thirteenth century a.p. that European writers began to
describe the games played in their respective countries. The first and most
important of these works is a manuscript (‘Alf’) which was compiled by
direction of King Alfonso X of Castile in 1283 and deals at length with the
games played in Spain.' For England, we have a short treatise on tables (‘K.’)
in MS. Brit. Mus. Kings 13 A. xviii. For Italy and France, we have the two
great collections of problems of chess, tables, and merels which are known by
' This work has been published in facsimile by Hiersemann of Leipzig with the title Das
Spanische Schachzabelbuch des Konigs Alfons des Weises von 1283 (Leipzig 1913). In addition to chess
and its derivative games, Alf. deals with gambling games, tables, and alquerque.
 INTRODUCTORY 3
the pseudonyms adopted by their compilers, Bonus Socius and Civis Bononiae.
Finally, for the Middle Ages we have a few smaller references in other manu-
scripts of which the most important is the Vetula, attributed to Richard de
Fournivall, Chancellor of Amiens in the thirteenth century.’ It is interesting
to note that the author thought it necessary to apologize for dealing with
such trivialities as games:
Sunt alii ludi parvi quos scire puellas
Esse decens dixi: sed parva movere pudebat
Nuncque magis, quam tunc, pudet illa minora referre.?
In the sixteenth century, Hierome Cardan (1501-76) wrote a work now
lost, De ludis latrunculorum, from which Hyde (p. 40) quoted a few passages
dealing with the Italian games of tables, and Rabelais (1. xxii) includes board-
games in the long list of games played by Gargantua. But it was not until the
middle of the seventeenth century that manuals of the current European
games began to appear. The Maison des Jeux académiques? and the Compleat
Gamester,* were the forerunners of many similar works in the various Euro-
pean countries. These are valuable, though they must be used with caution
since they tend to be encyclopedic, and, being generally based on French
manuals, often include games other than those played in the country of
publication. Willard Fiske made much use of these in his Chess in Iceland,5
and I have generally used his summaries of their contents.
The first scholar to deal with the non-European games was Thomas Hyde
(1636-1703), Professor of Arabic and Bodley’s Librarian, Oxford, in his De
Ludis Orientalibus,® a work which still remains of paramount importance.
Hyde, who was employed by the Crown to translate diplomatic correspon-
dence from Asiatic countries and met most of the envoys from these lands,
obtained his information on Oriental games first-hand. Our second source
for non-European games is the casual mention of games in early records
of travel. These are disappointing: early travellers were rarely interested in
games, and it is only incidentally that they are mentioned. Sometimes they
only give the native name of the game, but more often they call it ‘chess’ if
it looked complicated, ‘draughts’ if it looked simpler, or dismiss it as too
difficult to understand. It is only in the last sixty years, and largely due to
Sir E. B. Tylor and Dr. A. C. Haddon in Britain and Stewart Culin in
America, that the importance of board-games as a cultural phenomenon has
t An imitation of Ovid’s Ars Amatoria which was supposed to have been found in Ovid’s tomb
400 years after his death. Manuscripts are fairly numerous; there are two in the British Museum
(Harl. 3353 and 5263) and the work was printed in Cologne c. 1470 and 1479, and elsewhere.
I have used Wolfenbiittel 1661. It was translated into French in the fifteenth century (La Vieille,
ed. Vieille) Cocheris, Paris 1861.
2 Vetula, 1, xxxiv. 3 Paris 1657.
+ By Charles Cotton, London 1674.
5 Florence 1906. The title of this work is deceptive, since by far the larger part of the book
deals, not with chess, but with draughts, merels, fox and geese, and tables, for which it is a mine
of information.
® Oxford 1694.
4 INTRODUCTORY
been recognized, and competent accounts of them began to be a regular
feature in works dealing with the life and culture of native peoples.
These sources show that board-games are now very widely distributed and
are by no means confined to the more civilized peoples. The difficulty of a
game is no criterion of the development of the race playing it: some of the
simplest games are played by the more, and some of the most complicated
by the less civilized races. It may be that the only peoples without a board-
game are the Eskimos and the native tribes of New Guinea and Australia,
where the only games that have been observed are string-games. But there
are gaps in our information. We know very little of the board-games played
in Siberia and central Asia in general, in Morocco and Algeria, in the Bahr
el-Ghazal and the Congo basin, or by the native tribes of South America.
There remains accordingly a wide field for future research.
Game-manuals usually divide board-games into games of chance! and
games of skill. This classification, as we shall see in Chapter 2, goes back to
the ancient Greeks, but it is not sufficiently detailed, when we have to deal
with the large number of games that are included in this book. I have
accordingly adopted a classification based on the view that games are typical
of the early activities and occupations of man—the battle, the siege or hunt,
the race, alinement, arrangement, and counting.’I deal with the following
classes of board-games in Chapters 3-8.
I. GAMES OF ALINEMENT AND CONFIGURATION (Chapter 3). In the
former the players endeavour to aline a given number of their men: the
typical European games of this class are merels (three in a row) and go-bang
or pegity (five in a row). In the latter the players start from a particular
arrangement of their men and endeavour to rearrange them in another
configuration. The modern European games of halma and salta are of this
type.
2. WAR-GAMES (Chapter 4). These are of four types: (a) battle-games in
which the players endeavour to capture or immobilize all their opponent’s
men. The typical European games of this type are chess and draughts:
(b) struggles for territory in which each player endeavours to obtain control
of the larger portion of the board. The typical game of this type is the Chinese
wei-k’i (Japanese i-go): (c) blockade games in which no captures can be made
and the players endeavour to immobilize their opponent’s men: (d) clearance-
games in which the only moves are captures and each player endeavours to
make the larger number (or value) of captures. The typical English game of
this type is leap-frog.
3. HUNT-GAMES (Chapter 5) in which one player with the larger number
of men endeavours to hem in and immobilize his opponent’s men. The
typical European game of this class is fox and geese.
4. RACE-GAMES (Chapter 6) in which teams of equal size race one another
along a given track, and the first player to complete the course with his team
™ Some manuals subdivide these into games of pure chance and games of chance and skill.
 INTRODUCTORY $
wins. The moves are controlled by the throws of dice or other implements of
chance such as nuts, knuckle-bones, staves, or cowries, which I follow Tylor
in including under the term ‘lots’. The typical English games of this class are
backgammon, ludo, and the game of goose.
5. MANCALA GAMES (Chapters 7 and 8) in which the board consists of
two, three, or four rows of cup-shaped depressions or holes, each of which is
large enough to contain a number of beans, and the method of play is to lift
the beans from a hole and deal or sow them one by one in the following
holes. These games are widely spread in southern Asia, Africa, and the adja-
cent islands and were carried to America by African slaves.
Games of these classes are very unequally distributed. It is only in southern
Asia that games of all five classes occur. Europe has no mancala games.
Africa, apart from Egypt, Abyssinia, and those parts to which Muhamma-
danism has penetrated, has no war-games, no hunt-games, no race-games, and
no games of alinement. America, before the time of Columbus, Cortes, and
Pizarro, had only war- and race-games.
All board-games are limited as to time and space.! In time, they have a
definite beginning and a definite ending: in space, the action takes place
within a definite area, that of the board, which is marked off beforehand,
either materially or ideally, within which an absolute and prescribed order
must be observed. The men can only stand on the points or on the cells of
the board. Mathematically, all board-games are played on points, as may be
seen when chess is played on an in statu quo board, or when a latticed board
is set out on the ground in the form of an arrangement of holes. The use of
cells is convenient when the men are of some size, and minimizes the possi-
bility of dispute as to the exact location of a particular man. It would, how-
ever, be wrong to make a distinction between games played on cells and
games played on points, for while all games played on cells can be played on
points, the converse is not true. In early times players seem to have used
points or cells without preference for either. Chinese chess was originally
played on the cells but is now played on the points; in Europe, the Norsemen
used the points of an even-celled board and the cells of an odd-celled board
when they played tafl.
Games of the first three classes are games of manceuvre and require room
for manceuvring. As a rule, the outline of their boards is a square, but rect-
angles and triangular or circular boards also exist, and the square boards used
in the Asiatic war- and hunt-games are extended by the addition of small
triangles which are attached to the outer edges. Square boards are of two
types, depending on the way in which their interior is divided:
1. The interior bears a symmetrical arrangement of lines, their intersections
forming the points on which the men can be placed, while the lines give the
directions in which the men can be moved. The simplest of these boards is
the square with its orthogonal medians, giving nine points, which is used for
1 J. Huizinga, Homo Ludens, London 1949, pp. 9 ff., argues that this is true of all kinds of play.
 6 INTRODUCTORY
s nine holes. By the addition of the major diagonals, we obtain the board (still
of nine points) on which the smaller merels is played. The use of these dia-
grams was not confined to board-games: the first (with a circular outline)
os was used in the Minoan script as the symbol for 100! and the second (with
circular outline) occurs as decoration on the flat base of a saucer from Troy.?
Larger diagrams, from 20 in. to 5 ft. in diameter, are to be seen on the pave-
ments in the Forum at Rome} which from their size may have been used in
[‘,Ne games resembling hopscotch (Fr. marelle). The second diagram (with square
outline) has been long used in Ceylon as a charm against planetary and
\ demoniacal influence.*
The square, complete with diagonals and medians, also served as a unit of
design or to produce other game-boards. By combining two of these units,
we obtain the rectangle of 42 cells which heads a runic inscription at
St. Michael’s, Brough under Stainmore, Westmorland, attributed to the
period A.D. 550-600.5 Four units, arranged to give a square of 4X4 cells,
give the alquerque board (Fig. 27). Two alquerque boards, combined to
form a rectangle of 4x 8 cells, give the board of the Malagasy game, fanorona
(Fig. 36), and four alquerque boards, arranged to form a square of 8 x 8 cells,
give a board which already occurs on the roofing-slabs of the temple at
Kurna (see p. 19) and is now used for Indian and Indonesian games. Five of
the original units, arranged to form a cruciform board (Fig. 47), give the
board for the older fox and geese.
Another group of lined boards is obtained by using concentric squares in
which the middle points of corresponding sides of the squares are joined by
lines. Two such concentric squares form the board for the five men’s morris
(Fig. 18£) and three such squares give the older board for the larger merels
(Fig. 18F); in the latter board the corresponding angles are sometimes also
joined by lines (Fig. 18¢). The larger merels board (without the diagonal
; lines) also occurs at Kurna with a detached guarded cross in the central
\ z 5
\ » \. space, and is used in Ceylon as a charm? as well as a game-board.
2. The board is a square or rectangle, divided into a number of equal-sized
cells by parallels to the sides and making a lattice. In most games the men are
placed on the cells and no cell may contain more than one man, and the men
move from cell to cell in a prescribed way. In a few games the men are
placed on the points of intersection of the parallels and move along the lines
of the diagram. Latticed boards are generally squares, but the roughly
marked boards on stone that have been found in Palestine and on Roman
sites in Britain often show small differences in row and column, varying
from a norm of 12x 12 cells in Palestine and 8 x 8 cells in Britain, probably
through carelessness on the part of the players who made them. Presumably
™ Mosso, Palaces of Crete, London 1907, p. 71.
? Schliemann, Troy, London 1875, p. 284.
3 E, Falkener, Games ancient and oriental, London 1892, 364.
—* Parker, Ancient Ceylon, London 1909, p. $79.
5 Proc. Cumb. and Westm. Archaeological Soc. v (1881), 291. ® Parker, 579.
 INTRODUCTORY 7
in these cases the players fitted the number of pieces to the dimensions of the
board. But there are greater differences in dimension in the boards for some
games which must be deliberate, an increase in size with a corresponding
increase in the number of men to add to the length or difficulty of the game,
or a reduction in size and in the number of men to shorten the length of the
game or to reduce its difficulty. Thus, the draughtsboard, originally of 8 x 8
cells has been increased to 10X 10 cells for Polish draughts (4.3.7) and 12x 12
cells for Canadian draughts (4.3.8.),! and the board for the Norse game, tafl
(4.1.13), originally 18x 18 cells, was reduced to 11x11 cells (Wales), 9x9
(Lapland), and 77 (Ireland).
An existing game-board has often suggested or been used for a new game.
The inventor of chess arranged his game on the ashtapada board, and Indian
chessboards still retain features which belong to ashtapada and play no part in
chess.? The chequered chessboard has led to the transfer of alquerque and fox
and geese from lined boards to the cells of one colour on the chessboard.
Rithmomachy was played on a double chessboard. By doubling the number
of cells in row and column of the early Chinese chessboard of 9 x9 cells, the
board of 19 X 19 points was obtained and used first for an enlarged variety
of chess and later for wei-k’i. In more modern times, a number of Japanese
games have been placed on the wei-k’i board.
Primitive peoples, when playing board-games, use for the undifferentiated
pieces or men small objects that can be obtained in sufficient number, such as
pebbles, bits of stone, brick, or pottery, shells, seeds or beans, and even
pellets of camel, goat, or sheep dung. But kings and nobles in ancient Ur and
Egypt initiated at an early date the custom of using fashioned men, made
specially for use on elaborate and permanent boards. These men take various
forms, tall cylinders endirig in a knob or the head of an animal or man,
shorter cylinders with a hemispherical top, reels, circular disks with the
upper surface plane or slightly curved like buttons, hemispherical punt-
shaped pieces of wood, bone, or ivory, or pawns resembling the men we use
in halma or as chess-pawns. The use of circular disks with both faces plane,
which can be doubled or piled, seems to have originated in Greece or Rome,
and to have led to a modification of rule in the Egyptian race-game after its
introduction into Europe.
With board and men the materials for most games are complete, but all
race-games require also those implements of chance which I include under
the term ‘lots’. Lots are very old and were probably used for divination and
for gambling long before they were used for games. Presumably the earliest
tools that were used for divination were objects which, when thrown, could
only come to rest in two different ways; the vibhitaka nut used in India in
the Vedic period,’ the arrow mentioned in the Qu’ran (v. 92), in which. the
1 Similar increases in the size of the chessboard have often been made or proposed.
2 HC. 40-42.
3-H. Liiders, Das Wiirfelspiel im alten Indien, Berlin 1907, p. 17.
8 INTRODUCTORY
American anthropologists, F. H. Cushing and Stewart Culin' saw the original
of all lots, the stave or short stick, shaped by hand so that one face is plane
and the other curved, and the cowrie shell. The tetrahedral dice found in the
royal graves at Ur and in Palestine are virtually of this kind, since the four
corners of each die are cut away and two of them are marked by dots and
two are left blank.
It was not long before several of these objects were thrown together to
give a larger number of different values to the throws. Five staves were
found with a game-board in Tutankhamen’s tomb, and three tetrahedral dice
with a game-board in the royal graves at Ur. Six or seven cowries are used in
the Indian pachisi (6.5.6) and three or four staves in the American Indian race-
games. It was, however, soon discovered that the knuckle-bone of the sheep,
which when thrown could fall in four different ways, could be used instead
of three staves, and the knuckle-bone, or imitations known as astragals, came
into use and finally led to the use of the long die, a parallelopiped with square
ends and four longer rectangular faces, each of which could be given its own
value. Knuckle-bones have been found with game-boards in Egyptian tombs
and astragals in Greece. The long die, pasaka or pasa, already occurs in the
Epic Period in India, and still remains the favourite Indian form of die, and
sets of three or four pasas, loosely joined by a spindle, are also used. The values
of the faces vary: in the Epic Period in India they were numbered 1, 2, 3, 4,
but those used for the dice-chess had 2, 3, 4, 6, and those used for chausar 1, 2,
4, 6. The Roman tali had 1, 3, 4, 6, and two dice found in the lake-village at
Meare have 3, 4, 5, 6.
The last step in the story of dice was the invention of the cubic die. Hero-
dotus (i. 94) attributes this to the Lydians of Asia Minor, and is probably
right in so doing, for no cubic dice have been found in Egypt before the
seventh century B.c. and most belong to the Roman period. The cubic die
only reached India in the sixth century A.D. and has never taken the place of
the pasa. It reached China in the seventh or eighth century a.p. with nard,
the Persian name for the Greek tablé. Cubic dice from Greece of the fifth and
sixth centuries B.c. are in the Ashmolean Museum, Oxford, and many cubic
dice have been found in Etruscan tombs. The numbering of the faces of these
dice seems to have been early standardized, since the rule that the pips on
opposite faces must total seven is already adopted in the oldest Greek dice.
Etruscan dice, however, show other numberings in addition to the normal
arrangement. I have noticed 1 and 2, 3 and 4, 5 and 6; 1 and 2, 3 and 5, 4 and
6; 1 and 3, 2 and 4, 5 and 6. Occasionally early dice (e.g. three dice from Boge
in Jutland, now in the British Museum), may repeat the same number on
more than one face. This does not necessarily imply that they were used
fraudulently.
More complicated lots have been used at times. The British Museum has
Roman ‘dice’ with fourteen and twenty faces. The Alfonso MS. uses a seven-
' Chess and Playing Cards, Washington 1896, p. 679.
 INTRODUCTORY 9
faced die in some of its games. The English ‘long Lawrence’ was an eight-
sided prism. As a rule, however, when a larger selection of numbers is
required than that provided by the cubic die, use was made of a top or
teetotum, even in Roman times.
I conclude this chapter with a list of the powers of move and methods of
capture, which are exemplified in the games of alinement, war, and hunt in
Chapters 3, 4, and 5, in order to define the terms which I shall use when
describing these games."
MOVES
(a) IN THE OLD wor LD. The powers of move in the oldest games are
very simple and are related to the type of board used. When the game is
played on a lined board, a man can move one step only along a marked line,
which passes through the point on which the man stands, to the next adjacent
point on the line, provided it is empty. When the game is played on a
latticed board, a man can move orthogonally any distance, provided the cells
passed over are empty and the cell to which it is moved is also empty. This
is the rook’s move in chess.
In Europe during the Middle Ages, we find various tendencies at work:
(1) to omit the diagonal lines on lined boards with a consequent disuse of all
diagonal moves; (2) to abandon all powers of move in the backward direc-
tions;? and (3) to transfer war- and hunt-games on lined boards to the
chequered chessboards and to use only cells of one colour, with the result
that diagonal moves, previously restricted to some of the pieces in chess,
were introduced on latticed boards.
(b) IN THE NEW WORLD. Here our information about the native war-
games relates to two games only. In both games the ordinary power of move
is a single step in the diagonal directions and forwards only. In one game
(awithlaknakwe), where the first man to be taken is replaced by a seventh
man, this man, in addition to the ordinary forward diagonal move of one
step, can make a move of one step in the forward orthogonal direction.
We have, accordingly, the following powers of move:
GAMES PLAYED ON THE POINTS OF LINED BOARDS. A man posted on
any point of the board can only move along a line marked through that point:
(a) one step only, the point reached being empty;
(b) any number of steps along a straight marked line, provided all the
points passed over and the point reached are empty.
GAMES PLAYED ON THE CELLS OF LATTICED BOARDS. A man posted
on a cell can move to a vacant cell:
(a) orthogonally only, one step in any of the four possible directions;
™ I omit the moves which are peculiar to chess, and those used in rithmomachy which have
been borrowed from chess.
2 In compensation for this restriction of move, a man that reaches the opposite edge of the
board, and would then become immobile, is given promotion, with a new power of move so that
it becomes again mobile.
10 INTRODUCTORY
(b) orthogonally only, any number of steps in any of the four possible
directions, provided all the cells passed over are empty (this is the
rook’s move in chess);
(c) orthogonally only, one step forwards and sidewards but not back-
wards;
(d) orthogonally only, any number of steps forwards or sidewards, pro-
vided all the cells passed over are empty;
(e) diagonally only, one step forward in either of the forward directions;
(f) diagonally only, any number of steps in any of the four possible direc-
tions, provided all the cells passed over are empty (this is the bishop’s
move in chess);
(g) diagonally only, one step in any of the four possible directions, or if
this cell is occupied by a man of either colour and the first cell beyond
it is empty, the move may be a leap to this cell. If after this leap another
similar leap is possible, it may be made in the same turn of play, and
similar leaps may be made as long as they are possible;
(h) both orthogonally and diagonally, one step in the forward directions
only;
(i) both orthogonally and diagonally, one step in any direction, or if this
cell is occupied by a man of either colour and the first cell beyond it is
empty, a leap to this cell without capture, and the leaps may be con-
tinued in the same turn of play by further similar leaps as long as they
are possible.

CAPTURES
Board-games exhibit a greater variety in the methods of capture than in
the powers of move, but in the oldest games only two methods were used,
interception in games played on the cells of latticed boards, and a short leap
in games on lined boards.
In describing the methods of capture, I use “A’ for the player whose turn it
is to play and his men, “B’ for his opponent and his men, and ‘O’ for a vacant
cell. Thus, ‘ABO’ means that three consecutive points or cells in a straight
line in a direction in which a move can be made are occupied by (1) one of
A’s men, (2) one of B’s men, and (3) the third point or cell is empty, it being
A’s turn to play.
(a2) REPLACEMENT. A by a legal move can reach a point or cell occupied
by B. A removes B from the board and occupies its place. This is the capture
used in chess and in all race-games.
(b) INTERCEPTION. Three consecutive cells in a direction in which a
move can be played are ABO, and A by a legal move (not necessarily along
the row or column ABO) plays a man to O, producing the position ABA;
B is then taken and removed from the board, leaving the position AOA.
(c) LINE INTERCEPTION. A number of consecutive cells along a row,
column or slant line of cells (all of the same colour on a chequered board) is
 INTRODUCTORY II
ABB...BO, and A enters a man on O, producing the line ABB...BA.
A turns over all the intervening Bs, and the line is now an unbroken sequence
of As. If other lines through O are similarly occupied by an unbroken
sequence of Bs, ending in A, A in the same move turns over all the Bs in
each of these lines.
(d) INTERVENTION. Three consecutive cells in a direction in which a
move can be made are BOB, and A by a legal move places a man on O,
producing the position BAB. Both Bs are taken and removed from the
board, leaving the position OAO.
(e) SHORT LEAP. Three consecutive points or cells in a direction in which
a move can be made are ABO. A leaps over B and occupies O, removing B
from the board and leaving the position OOA. If the rules of the game end
the turn of play after this first capture, I term the capture single. If, however,
the rules permit a series of similar captures in a turn of play, I term the capture
multiple.%
(f) Lone LEAP. This is single if the rules only permit one capture on a
move, and multiple if a succession of captures are permitted in a turn of play.
If A and B lie on points or cells in a direction in which A may move, and are
separated and followed by a series of empty points or cells so that the position
is AOO...OBOO...O, A leaps over B and occupies one of the series of
empty points beyond B.
(g) LINE LEAP. A series of consecutive points in the direction of move is
ABB...BO, the number of Bs being odd. A leaps over the unbroken series of
Bs and occupies O, the row of Bs being taken and removed from the board.
(h) apPpROACH. A series of points along a straight marked line on the
board is AOBB...B. A moves to O and takes the unbroken series of Bs and
removes them from the board, the final position being OAOO . . . O. Ifnow
A can make a capture by approach or withdrawal (see i) along another line
of the board, he proceeds to do so until no further captures in this way can
be made.
(i) WITHDRAWAL. A series of points along a straight marked line on the
board is OABB...B. A withdraws to O and takes the unbroken line of Bs,
producing the position AOOO...O. If now further captures, either by
approach or withdrawal, are possible, the player makes them in the same
turn of play.
(j) HUFE. In some games in which captures are compulsory, and a player
neglects to make a possible capture, making another move, his opponent
removes one of the player’s men as a penalty for his default.
(k) Inthe larger merels, a player, who succeeds in alining three of his men on
consecutive points along a line of the board, takes one of his opponent’s men.
In some games the captured man may not be one of those forming a row.
1 In some games in which multiple short leaps are permitted, none of the captured men can be
removed until all the leaps have been made and the leaping man has come to rest. In these cases
the captured (but unremoved) men exercise the same restraint on A’s liberty to move as if they
had not been taken.
 Z
GAMES INTHE ANCIENT eWORED
Tue recorded story of the ‘drama of history’ begins with the early civiliza-
tions of ancient Egypt and Babylonia, and in both lands the existence of
board-games is one of the earliest features of their civilizations. There is also
evidence that these or similar games were played in Palestine and Assyria,
in Cyprus and Crete in the second millennium B.c., and in Greece and Rome
in the first millennium B.c. The evidence for this is summarized, and the
natures and relationships of these games are discussed in this chapter.
The earliest-known example of a gaming-board with accompanying
pieces was exhibited in July 1909 at King’s College, London, by the Egyptian
Exploration Fund, and is now in the Musée du Cinquantenaire, Brussels. It
was discovered by officers of the Fund when excavating a predynastic
cemetery at El-Mahasna, about eight miles to the north of Abydos in Upper
Egypt. The catalogue of the exhibition gave the following account of the
grave (H. 41) in which it was found:
A large grave containing six burials, which had been partially plundered. At the
south end of the grave were standing eight pottery vases, and the four clay objects
were found with them. These are hollow and pierced for suspension, and when shaken
make a noise like a rattle. They may have been part of the outfit of a medicine man, or
may have been put to the more prosaic use of goat or sheep bells. The ivory comb was
also found at this end of the grave. ... Bunches of imitation garlic and a large number
of beads were found. At the other end of the grave was found a clay gaming-board.
This is undoubtedly the origin of the sign ‘men’ (#4) in the hieroglyphs. Ivory brace-
lets, clay objects of unknown use, and a fragment of a clay wand complete this
group."

The board is about 7 in. long and 23 in. wide, and stands upon two low
cross-pieces. The surface is divided into 3 x6 cells. Eleven conical pieces were
found with it, two about 1 in. in height and base-diameter, and nine others
about $ in. in height and base-diameter.
The predynastic period in Egypt extended from about 4000 to 3500 B.c.,
so this board may be a thousand years older than the next evidence of
gaming-boards. In view of the surmise of the explorers that the grave might
_ be that of a medicine man or magician, we cannot say for certain whether
this board was used for a game or not. But it has definite affinities to later
boards that have also been found in tombs. Like these, it has three rows of
cells and is accompanied by conical pieces. Whether it was used for divination
? See also E. R. Ayrton and W. L. S. Loat, Egyptian Exploration Fund, 1911.
2urer| GAMES IN THE ANCIENT WORLD 13
only or for a game, it shows that the craftsman who modelled it and the
pieces had already acquired the idea of moving men on a pattern of cells,
which is fundamental to all board-games.

FIG. I. PRE-DYNASTIC BOARD FROM EL~MAHASNA

1. Ecypt
WALL-PAINTINGS. Paintings begin to appear on the walls of temples,
burial-chambers, and other buildings of the Fifth Dynasty (c. 2500-2400 B.c.),
and these sometimes include pictures of persons playing board-games with
accompanying text. One of the oldest of these paintings is in the tomb of
Ra-sheps (Rashepses), a scribe or functionary of the court.! It shows two
games in progress:
2.1.1. Sen’t or senat. Two men are kneeling at opposite ends of a board
which shows an unbroken line of twelve men, cylindrical in shape but the
alternate men have knobbed tops. Each player is touching one of his men,
and an inscription above the board says ‘see the sen’t’ and above each player
‘takes three from the board’.
2.1.2. Han, ‘bowl’ or ‘vase’. Two men are playing a game on a board
that consists of twelve concentric circles, and two men are looking on. The
inscription says ‘the game of the bowl’.
A similar painting in a tomb at Beni-hassan of the Twelfth Dynasty
(2000-1780 B.C.) also shows two games. Neither is named, but the first is
identical with the sen’t of Ra-sheps’ tomb. The second is probably
2.1.3. T’au, ‘robbers’. Two men are kneeling at the ends of a board on
which there are five pieces at each end with an empty space, sufficient to hold
three or four men, between the two armies.
An inscription at Thebes? gives the names of three games, han, sen’t, and
t’au; and Dr. S. Birch} suggested that the second game at Beni-hassan is f’au.
' C.R. Lepsius, Denkmadler, ii. 61; Tomb 16.
2 H. K. Brugsch, Monuments Egyptiens, pl. lxviii f, h; J. F. Champollion, Notices descriptives,
566.
3 Revue Archéologique, xii (1865), p. 65.
14 GAMES IN THE ANCIENT WORLD [GH
Pictures of t’au also occur in the various recensions of the Book of the Dead,
where the soul of the departed plays a game in the other world, and a painting
at Medinat-Habu shows Rameses III (Twentieth Dynasty, c. 1270 B.C.) play-
ing with Isis. The latest picture of this game occurs in a well-known papyrus
manuscript, now in the British Museum, which shows a game between a
lion and a goat, each with four men on the board while the goat holds one
man in his hand and the lion a man in one hand and a bag (presumably to
contain the men) in the other.

FIG, 2. FROM TOMBS AT BENIHASSAN

These pictures show that the ancient Egyptians played at least three board-
games, but little more, and a Turin manuscript with an account of one of
the games, which has been translated by A. Wiedemann,' gives little help.
The drawings are probably conventionalized but the arrangement of the
men suggests, first, that sen’t was a race-game, although there is nothing in
the drawings to show any use of dice or lots, and second, that t’au was a
war-game. We know nothing about han, and Falkener’s attempt (pp. 83-90)
to reconstruct the game is only a guess and quite unconvincing.? That f’au
was a war-game seems to be supported by the “Books of the Dead’: a war-
game which depends on skill alone seems more suitable for play in the other
world than a game of chance.
GAME-MATERIALS. Game-boards, men for games, and lots have been
found in tombs from the First Dynasty onwards, and frequently occur
together among the treasures buried with kings and nobles from the Twelfth
Dynasty (c. 2000-1788 B.c.) onwards.
2.1.4. A middle-class cemetery dating from the First Dynasty (c. 3000 B.C.)
was found in 1936 during a re-excavation of the Avenue of Sphinxes at
* ‘Das Brettspiel bei den alten Aegypten’ in Actes du dixiéme Congres d’Orientalistes, Leiden 1897.
? A similar board to that of han is used by Yoruba children at Nopa, Southern Nigeria,
for a simple race-game which is known as ashere ‘playing’. The board consists of seven con-
centric rings, and four or five children sit round the board, each having a single man which
is entered in the outer ring and moves one ring at a time until it reaches the middle ring
where it is borne, and the first player to bear his man wins. Each player in turn conceals a
bit of stick in one hand and offers both hands to the next player who chooses one hand. If it
contains the bit of stick, he enters his man in the outer ring, or advances it one ring and the
turn of play passes to him. If he chooses the empty hand, the first player advances his man
one ring and his turn of play continues until he loses the bit of stick. K. C. Murray who saw
the game in April 1951 says that there is a song associated with the game which the children
sing as they play, and that the board sometimes consists of seven concentric squares.
Sees
2.1.4] GAMES IN THE ANCIENT WORLD 1§
Sakkara. One undisturbed grave yielded a set of game-pieces, seven cylindri-
cal in form and about 8 cm. high, and seven hemispherical and about 4 cm.
in diameter.!
2.1.5. A game, name unknown but generally called dogs and jackals by
Egyptologists,? appears in tombs from the Ninth or Tenth to the Twelfth
Dynasties, and spread to Palestine, Sumer, Elam, and Assyria. It must have
been very popular for a time from the number of boards and fragments of

G’ 24 23 22 2] F’ 19 18 17 16 E 4 13:12 “I
OW) 0 £0'0'507'@)_010
7020" @.0n0n Ome

%
A
O

. 04.0. C0705 OOS) 207.0100

6 G 24.25.22.2) 14 13
F 19.18 I7 IGE fou)
FIG. 3. EGYPTIAN RACE GAME. XI AND XII DYNASTIES

boards that have survived. It was obviously a race-game. The board is

roughly rectangular, but one end of the Egyptian boards is curved and the
sides are sometimes curved to resemble an axe-blade. The points of the track
are marked on all the boards by holes for the use of pegged men, and ten
carved ivory pins, half with heads of dogs and half with heads of jackals,
were found with one board from Thebes. The general layout of the boards
is shown in Fig. 3. Two persons played, one entering his men at A and follow-
ing the track through B, C, &c., to H, and the other entering at A’ and follow-
ing the track through B’, C’, &c., to H, each track containing thirty points (or
perhaps both continue after H, the first ending at A’ and his opponent at A).
The points marked by letters in Fig. 3 are distinguished in some way on the
Egyptian boards, by mark or inlay or by circles. Points 6 and 20, 8 and 10
are connected by lines which may have served the same purpose as the
ladders in our snakes and ladders, a man reaching 6 being transferred at once
to 20 and so on. From the fact that four points lie between consecutive points
with special markings (omitting C and C’), I infer that the moves were given
by the throw of four staves (or some other form of lots) which give five
different throws.
ESI; 2) Jan. 1937:
2 Sir Flinders Petrie (Objects of early Use, $3) called it the Game of 58 Holes, but the number of
holes varies on different boards, and in Lord Carnarvon and H. Carter, Five Years’ Excavation at
Thebes, 56 ff., it is called Hounds and Jackals from the pieces found with a board from Thebes.
16 GAMES IN THE ANCIENT WORLD [CH. 2
The oldest of the dated Egyptian boards (Ninth or Tenth Dynasty) was
found at Sedment.! Next comes an Eleventh-Dynasty board, now in the
Metropolitan Museum, New York, which was found by H. E. Winlock at
Deir el-Bahri in 1925-6. This board had been so frequently used that the
holes on one edge had been worn out and a patch put in to replace them.?
A finer ivory board of the Twelfth Dynasty, now in the same museum, was
found by Lord Carnarvon at Thebes; it was accompanied by the game-pieces
mentioned above.3 Another board of the Twelfth Dynasty from el-Lahun
is described in Petrie.* In addition to these boards, Sir Flinders Petrie5 refers
to two more in the Cairo Museum, and there is one in the Constantinople
Museum, Eski Sark, no. 4065.°
The ‘Feudal Age’ in Egypt came to an end with the Hyksos invasion,
c. 1680 B.c., and materials of their hundred years’ domination are very
scanty. What is known as the ‘Empire Age’ followed, c. 1580 B.c., and the
Eighteenth Dynasty, which united all Egypt under one rule, opened a period
of great prosperity, marked by an efflorescence in all kinds of art and architec-
ture. Many tombs of this and the next two dynasties have been opened which
contain game-materials. These usually take the form of a long narrow box
which is often accompanied by a long stool on which the board can be
raised. The box contains a drawer for the men and lots, and both surfaces
are marked for the boards on which the games were played.
2.1.6. One side of the box bears a board of twenty cells; at one end three
rows each of four cells and the middle row is continued along the box to add

FIG. 4. EGYPTIAN GAME-BOX. UPPER FACE

eight cells. In some boards, for example, Queen Hatshepu’s board in the
Louvre,? the fourth, eighth, and twelfth cells, starting from the end with
three rows of cells, are specially marked.

1 W. M. Flinders Petrie and G. Brunton, Sedment, i, pls. xxi, xxii, and pp. 7, 12.
2H. R. Winlock, Excavations at Deir el-Bahri, 1942, pl. xxxvi and p. 129.
$ Carnarvon and Carter, op. cit., pl. land pp. 55f.
* Kahun, Gurob and Hawara, pl. xvi. | am indebted for these references to Mr. C. J. Gadd, ‘An
Egyptian Game in Assyria’, in Iraq, vol. i, part i (April 1934), pp. 45-So.
5 Sedment, i. 8, pl. xxii.
® G. Marin, ‘An Ancestor of the Game of Ludo’, Man, xlii (1940), 64. For the Asiatic boards
for this game, see pp. 21-23.
7 P. Pierret, Musée du Louvre, Paris 1873.
2.1.7] GAMES IN THE ANCIENT WORLD 17
2.1.7. The other side of the box bears a board of 3 x 10 cells. Here again,
some (but not all) boards have marked cells. Starting from one end of the
board, the first four cells of one of the outer rows are numbered 1, 2, 3, and
4, and the fifth cell bears the hieroglyph , meaning ‘gate’ or ‘door’. The
numbers suggest that this game was played with the help of knuckle-bones

a a) fe
fea | a
|
LILLIA IR IL
FIG. §. EGYPTIAN GAME-BOX. LOWER FACE

or some form of lots that give four different throws, and that the men were
entered on. the numbered cells and entered the main track by the ‘gate’.
2.1.8. A board from a tomb at Ak-hor, now in the Cairo Museum,
substitutes 2 board of 312 cells for the more usual 3 X10 board. In this
board the boundaries of the cells are of different widths; four lines separate

ASSES |S
T_T SNL
ees
|al eeisosee
FIG. 6. BOARD FROM AK-HOR, CAIRO MUSEUM

the rows, and the board is divided into two parts, one of 3 x 4, the other of
3X8 cells. All the cells of the 3 x4 are separated by four lines, but those of
the outer rows of the 3 x 8 are separated by two lines, and of the middle row
by three lines.! The first, fourth, and sixth of the middle row, starting from
the 3 x4 end, are marked with birds.
In all these boards the cells are either sunk or the bounding lines are raised.
This shows that the games were played on the cells and not on the points of
the boards. The relative sizes of the cells and of the men show that only one
man could occupy a cell at a time.
When these boards are found in an undisturbed grave, the drawer often
1 Are these differences a clue to the order in which the cells are traversed? Cf. the Roman
duodecim scripta board from Ostia, Fig. 16.
5356 Cc
18 GAMES IN THE ANCIENT WORLD [CHa
contains pieces. Generally there are two sets of men, one set of ten pawn-
shaped men and the other of ten reel-shaped men, both sets being half light
and half dark coloured. In one of Queen Hatshepu’s boards (now in the
British Museum) the drawer contained twenty lion-headed men, ten of light
coloured wood, nine of dark wood, and one of ivory, which perhaps replaced
a lost dark piece.
The drawers often also contain lots. Thus, four staves were found with
one board, and two knuckle-bones with another from Tutankhamen’s tomb,!
and a long die was found with the board from Ak-hor (2.1.8). All these
games are accordingly race-games in which the course of play is complicated
by the marked cells. The analogy of all race-games of the Old World makes
it probable that these cells were cells of safety and that a man stationed on a
marked cell was immune from capture.
The first of these games (2.1.6) appears to be the sen’t of the wall-paintings,
since both used a board twelve cells long. The second and third games
(2.1.7 and 8) are, I think, modifications of sen’t which bridge the gap between
sen't and the Greek kubeia and form part of the pedigree of backgammon. So
far, however, we have found nothing that is comparable with t’au (2.1.3)?
although Plato, as we shall see below, says that the Greek war-game was
introduced from Egypt. It is obvious that the wealthier classes which could
afford the magnificent tombs which have preserved these boards and men,
preferred the race-games, but there is evidence that other games were
played.
2.1.9. A stone board of 11x14 cells is in the Egyptian Galleries of the
British Museum (no. 56924), and also a tombstone group showing men play-
ing on a board composed of cells (no. 11888).
2.1.10. More important are the many diagrams of game-boards which are
incised on the great roofing slabs of the temple at Kurna on the western side
of the Nile at Thebes, the erection of which was begun by Rameses I (1400-
1366 B.c.) and completed by Seti I (1366-1333 B.c.), which Parker (pp. 579,
644) describes. These were all probably cut by the masons who were shaping
the slabs before they were finally placed in position on the roof, for three of
the diagrams were partly cut away when the edges of the slabs were trimmed
to make them fit against the adjoining slabs. Parker noted seven distinct types
of board which I reproduce in Fig. 7. All of these diagrams, except D (which
the mason seems to have left unfinished because he had made a mistake),
have been used as game-boards in different parts of the world.? It is difficult

ILN., 5 Oct. 1928.

? Falkener, plate facing p. 46, gives a board of 12x 12 cells which he had ‘restored’ from the
fragment of a board of the type of Fig. 6 from Queen Hatshepu’s tomb, now in the British
Museum (no. 21577). The board was undoubtedly a box-board of 3 x 10 cells in which the ‘gate’
was marked by the hieroglyph ‘nefer’. The restoration must be rejected entirely.
3 Ais the nine-holes board (see p. 39 and Fig. 188); B and C are the smaller merels board (see
p. 40 and Fig. 18D); D is probably intended for the alquerque board (Fig. 27); E is the pentagram,
used today in Crete, Spain, and by the American Indians for a game, and probably also in Ancient
2.1.10] GAMES IN THE ANCIENT WORLD 19
to see any reason for the cutting of these diagrams other than for use in
playing board-games.
Parker also found the cups of the mancala board incised on the roofing

2 NAANVASA
MYIMVIN
NZ
NINA)
INVAINNANIAIS
G

E
FIG. 7. GAME DIAGRAMS ON THE ROOFING SLABS OF THE TEMPLE AT KURNA
(after Parker)
slabs at Kurna, the tops of the walls of the temples at Karnak, Luxor, and
elsewhere, which he considered to be as old as the above diagrams. I shall
deal with these in Chapter 7.
2. UR!
2.2.1. Among the treasures discovered by Sir Leonard Woolley and his
exploration party, when the royal graves of the First Dynasty of Ur (after
c. 2560 B.C.) were opened in 1926-7, were four game-boards. These only
differed in the decoration of the cells. The boards are oblongs of 3 x 8 cells
from which two cells have been omitted on each of the outer edges. They
Greece (p. 28); F is the larger merels board (see Fig. 18); G is the quadruple alquerque board of
8 x 8 cells which is used today in Indonesian games (see p. 68 and Fig. 31).
T See also p. 36.
20 GAMES IN THE ANCIENT WORLD [CH.2
therefore contain twenty cells like the Egyptian boards (2.1.6), and, like
them, they contain a drawer to hold the men and lots. The upper surface of
the board is covered with bitumen in which little squares of shell are fixed to
form the cells. In the more elaborate boards the shell plaques are inlaid with
lapis lazuli and red limestone and divided by lapis-lazuli strips. In other
boards the cells are engraved with animal scenes. All the boards agree in
having rosettes on the middle row on one side of the bridge (where the cells

\iN
SS

TAM
il

FIG. 8. GAME-BOARD FROM THE ROYAL GRAVES, UR

are lacking in the outer rows) and on the two outer row cells on the other
side of the bridge; the more elaborate boards have also rosettes on the two
outer edge cells of the 3 x 4 end.
With the boards were found the men, seven counters on each side,! one
side white with five darker spots, the other side black with five lighter spots,
and two sets of three tetrahedral lots with the corners cut off and two of the
corners dotted with inlay.? A similar lot was found at Tell Beit Mirsim,
Palestine.
I think that there can be little doubt that this game is the same as the
Egyptian sen’t notwithstanding the difference in shape. Both games are
arranged for the use of a number of lots which give four different throws. In
the more elaborate boards of both games, the marked cells are the fourth,
eighth, twelfth, &c., starting from the inner end of the group of four cells on
the outer rows and passing down the middle row, and this indicates the
direction of the earlier moves. I suggest that the track in both the Ur and the
Egyptian boards consisted of twenty-seven points, but the turn was taken

* Cf. the Sakkara set of men (2.1.4). The Egyptian race-games of the Empire Age seem to have
generally required sides of ten men (p. 18).
? Sir L. Woolley, Ur, thefirst Phase (Penguin, 1946), p. 35 and pl. 11.
22.1] GAMES IN THE ANCIENT WORLD 21
differently, and the men were borne from different cells. In the Ur game the
men, on reaching the penultimate cell of the middle row, pass to a rosetted
cell on the outer row, pass round the outside of the 3 x 2 rectangle and enter
the middle row on the penultimate cell, then return down the middle row,
and finally to the outer row of the 3 x 4 rectangle from which they are borne
by appropriate throws of the lots. In the Egyptian games, the men on reach-
ing the end cell, return down the middle row and are borne from the last
four cells of this row. It will be seen that in each game, by playing in this
way, every fourth cell is marked with a rosette or marked in some other
way. I suggest also that each player had originally seven men, but that in the
Egyptian games the number was first reduced to six, as in the wall-paintings,
and finally increased to ten as in the boards of the Empire Age.
2.2.2. Three boards for the Egyptian game of ‘dogs and jackals’, have
been found in Mesopotamia, two of them at Ur. The first, a clay plaque,
4X 214 X jz in., was found by J.E. Taylor
during his excavations at Ur in 1853-4,
and is now in the British Museum (no.
123331) and was first published by C. J.
Gadd.! The board differs from the
boards in the Metropolitan Museum,
New York, in four ways: the marked
holes are surrounded by circles, the
holes 11-14 (Fig. 3) lie along the shorter
side of the board, the lines connecting FIG. 9. BOARD FROM UR
B and F, B’ and F’, C and D, C’ and (B.M. 123331)
D’, are omitted, and two extra holes
are inserted between A and A’ (Fig. 9). The second, now in the Uni-
versity Museum of Pennsylvania, was found by Sir L. Woolley at Ur
in 1931-2, illustrated on p. xiv of this book. It adds two holes to the track
before the beginnings of the track, and the ornamental group of figures, now
broken, attached to the short straight side, is characteristic of the Asiatic
boards. The board is 13 cm. in length. A third board is in the Baghdad
Museum (no. 17876). It differs from the second example in transferring the
rosettes on the opening holes to the new holes added in the Asiatic boards.
All these boards are considerably later than the boards from the royal tombs,
and Mr. Gadd thinks that it is by no means impossible that they should date
from the Assyrian rule over Ur.

3. AASSYRIA
A number of fragments of boards for the Egyptian “dogs and jackals’ are
described by C. J. Gadd.? They are all drawn on slabs of stone which had
been obtained from Egypt, and were made after the conquest of Egypt by
1 ‘An Egyptian Game in Assyria’, Iraq, i (1934), pl. viiib, and p. 46.
2 Op. cit., pl. viia, b, viiia, pp. 46-49.
22 GAMES IN THE ANCIENT WORLD [CH. 2
Esarhaddon, c. 675 B.c., for some of the boards have on the reverse side a
four-lined inscription:
Palace of Esarhaddon, the great king, the mighty king, king of all, king of
Assyria, governor of Babylon, king of Sumer and Akkad, son of Senna-
cherib, the mighty king, king of all, king of Assyria, son of Sargon, the
mighty king, &c.
The shorter edge of these boards either contains mortices by which groups
of figures (generally of animals) canbe added in the same plane as the board,
or fragments of such figures. The most important of these fragments are one
in the Constantinople Museum (no. 4646), and in the British Museum, nos.
123333 (found at Nineveh by Sir H. A. Layard), 12102 (an unfinished board),
91930, 12104, 81-7-27, 183, and 81-2-4, 19 (from Quyunjiq) and 118768
from Abu Habbah (Sippar). There is no evidence for the existence of this
game in Babylonia or Assyria before the time of Esarhaddon, though Elam
is represented by several fragments of boards, all found in the foundation-
deposits of a temple built at Susa by the King Shilhak-In-Shushinak who
reigned in the latter part of the twelfth century B.c. One of these has the
feature unique outside Egypt, of having on the reverse of the stone a board
marked out in squares for another kind of board.
A board, shaped like a mitre or shield, with three larger holes at the angles
and a number of smaller holes along the sides, was found when excavating
the site of a Hittite town at Boghaz-Kéi in Asia Minor dating from about
the thirteenth century B.c., which may be the board for a similar game."
4. PALESTINE
Game-materials, dating from the period c. 1700-1500 B.c. onwards, have
also been found in Palestine. R. A. S. Macalister? sums up the position thus:
At almost all stages of a Tell we may find flat boards of limestone, ruled in squares
by means of vertical and horizontal lines. It is notable that there is a very considerable
variety in the disposition of these squares. Sometimes we may find a perfect square of
twelve vertical and twelve horizontal rows of compartments. Sometimes there are
three horizontal and six vertical rows. Sometimes certain crossings of lines are marked
with X, but as we do not know the details of these ancient games, we cannot say
what this mark signifies. The ‘men’ used were either pebbles, or else small conical
stone pieces resembling those used in the game of halma.

Button-shaped men of excellent finish were also found at Samaria in 1933.

Three fragments of limestone boards, apparently of 12 x 12 cells were found
at Tell Zakariya.3 At Tell Beit Mersim, five cones, five tetrahedral lots in
blue faience, and one teetotum, shaped like a truncated pyramid, the sides
bearing one, two, three, and four dots, were found.
? Gadd, 46, citing K. Bittel, Klein-funde aus Boghaz-Koi, p. 23 and pl. 14; Revue d’ Assyriologie,
Xxxix. 30 ff.
? A Century of Excavation in Palestine, London 1925, p. 244.
3 PEFQS., April 1899, p. 99 and pl. 4 Ibid., July 1904, p. 215.
2.2.2] GAMES IN THE ANCIENT WORLD 23
There is also evidence that the Egyptian game of dogs and jackals was
known in Palestine. A plaque which was found at Gezer was claimed by
Prof. Macalister! as ‘a highly conventionalized re-
presentation of the Queen of Heaven’, but the
plaque contains a series of holes round the side
edges and up the middle which would be identical
with those of the Egyptian game-boards if only
the holes forming the ‘head’ were reduced to one.
Macalister thought that the holes were too small
and shallow to hold a pegged piece, but J. Gadd
points out that “several of these holes were sur-
rounded by a small circular spot of darker
red’.2 A fine ivory and gold object of the thir-
teenth century B.c. from Megiddo, which seems
to have belonged to this game, was described in
ILN3 FIG. I0. TERRA-COTTA
PLAQUE FROM GEZER
5 CYPRUS (after Macalister)
2.5.1. A board of twenty cells (see Fig. 11),
identical with the Egyptian board for sen’t, and with the same arrange-
ment of marked cells has been found in Cyprus.4

FIG. II. GAME-BOARD FROM ENKOMI (Cyprus)

6. CRETE
A very elaborate board, attributed to the period 2000-1700 B.c., was found
in the Minoan palace at Knossos, and is described and illustrated in the Annual
of the British School at Athens. Although no pieces were found with it, it was
at once assumed to be a game-board, and even a chessboard (as in the caption
on the postcards on sale at Knossos). The board is quite unlike any known
game-board in its arrangement, and the few cells, ten at one end,® do not in
my opinion suggest a game at all.
™ See Fig. 4 from R. A. S. Macalister, History of Civilization in Palestine, Cambridge 1921,
p. 91.
2 Gaad, p. 46, n. 2. 3 23 Oct. 1937, Pp. 709.
4 JHS., 1896, p. 288; and A. S. Murray, Excavations on Cyprus, p. 12 and fig. 19.
5 vii (1900-1), 77-82.
6 The four larger and four smaller circles at the other end do not appear to have any relation
to the ten.
 24 GAMES IN THE ANCIENT WORLD (Ost
Unmistakable game-materials have, however, been found in Crete. The
nine-holes board (Fig. 18a) is incised on a step of the theatre at Knossos,! and
4 }
—————
PAR SIDA DID SS la!
2 x x =)

Ur,
ONG
edLANG:

vy,
ROIAR
>)
+s =
Ne
Spccssxe7 ‘|S
Ne:a
va
ANY
AVS) SUZ SEIT STS, AT
DE GUAS DID
R
SOC SSE A
SKU USK N
STC SAAS ALI OS
SBS aS CUI
Ses |
FIG. 12. BOARD FROM KNOSSOS, CRETE

conical ivory men were found in a deposit dating from the Middle Minoan
Illa period (c. 1780-1580 B.c.) with prisms showing rows of circles and dots
on each side, which were probably lots.?

7. GREECE
The existence of board-games in Greece during the classical period is
attested by a small number of references to games in Greek literature from
Homer downwards. These references were collected and discussed by the
antiquaries of the second and following centuries A.p., but, as I have already
remarked, their efforts to explain them have only added to their obscurity.
All earlier attempts to elucidate these games have been superseded by a lucid
article by Professor R. G. Austin, “Greek Board-Games’ in Antiquity,3 and
what I have to say in the following paragraphs is based on this article. But I
have also endeavoured to deal constructively with the whole matter in the
light of what has already been said in this chapter about the games of the
near Orient, and to clear up some of the obscurities which Austin has left
unsolved. I believe that the board-games of Greece and Rome are affiliated
to the older games of Egypt, Ur, and Palestine, and that they reached Greece
by way of the Mediterranean Islands. This is the view first put forward by
Plato when (Phaedr. 274d) he assigns an Egyptian origin both to petteia and
kubeia.
T Mosso, 313. ? J. D. S. Pendlebury, Archaeolcgy of Crete, 1939, p. 167.
3 Sept. 1940, pp. 257-78. Ausf, 4
2.5.1] GAMES IN THE ANCIENT WORLD 25
The word petteia is derived from the Greek pessos, ‘pebble’, which was
also used in the sense of ‘game-piece’. It is a common phenomenon of the
nomenclature of board-games to give the game a name derived from the
name of the men with which they were played. The Romans used ludus
calculorum for any board-game played with calculi, ‘pebbles’ or ‘counters’, and
specifically Iudus latrunculorum or latrunculi for their war-game played with
latrunculi, ‘soldiers’. The same usage is seen in the Romanic languages of
medieval Europe. When chess, tables, and merels reached southern Europe
they were named respectively scacci, tabulae, and merelli because they were
played with scacci, “chessmen’, tabulae, ‘table-men’, and merelli, ‘counters’. In
the same way, draughts was named first ferses and later dames because it was
played with chess queens, and these were the current names of the chess
queen. This rule of nomenclature is not confined to Europe; both in Asia and
Africa mancala is often called by the name of the beans or seeds which are
used as pieces in this game. In naming their war-game petteia, the Greeks were
only following a common practice. In an analogous way they named their
race-game kubeia because the game made use of kuboi, ‘six-sided dice’.
I think it is clear from Plato’s references to games that to him petteia and
kubeia were particular games and not classes of games. They were not neces-
sarily the only games played in Athens, but they were the ones in which a
mastery of the game could only be achieved by long and continuous practice
(Rep. 374d). But it was an easy step in the course of time to use the words
generically, petteia for any game without, and kubeia for any game with dice,
and then to go farther and use either word for each kind of board-game, or
even for any kind of game, whether it required a board or not. Austin has
shown that all these steps had been taken before the antiquaries began to pay
attention to the games played by their forefathers. The first to do this were
two scholars of the first and second century A.D.—Suetonius, who wrote a
work on Greek games which is now lost except for a few extracts made by
Eustathius in the twelfth century,’ and Pollux, whose Onomasticon has sur-
vived. In this work (ix. 97-98), Pollux deals with a group of words derived
from pessos (which he says is a synonym for pséphos) and continues:
(97) Each of the players has five men on five lines, so that Sophocles naturally says
‘five-lined boards and the throws of the dice’; and of the five lines on either side there
was a middle one called ‘the sacred line’, and a player who moved a piece from it gave
rise to a proverb ‘he moves the piece from the sacred line’.
(98) The game played with many pieces is a board (plinthion) with spaces disposed
among lines: the board is called polis, ‘city’, and each piece kuon, ‘dog’; the pieces are
of two colours, and the art of the game consists in taking a piece of one colour by
enclosing it between two of the other colour. . . . Next to this game is diagrammismos,
a game which used to be called grammai, ‘lines’.?

' The Suetonian fragments are collected in A. Reifferscheid, C. Suetoni Tranquilli Reliquiae,
1860, pp. 322 ff.
2 Plinthion, as Austin points out, is also used by Arrian and Josephus of a column or mass of
 26 GAMES IN THE ANCIENT WORLD [CH
Hesychius (fifth century A.D.) calls diagrammismos ‘a game of sixty pieces,
white and black, moving on spaces’, and Eustathius repeats this with the
addition ‘a kind of kubeia’. Pollux, elsewhere in his work, includes diagram-
mismos in a rather miscellaneous catalogue of kubeia games, which includes
some games that are not board-games.
Pollux’s account of the early Greek games is the most explicit of those
given by the antiquaries. It is surprising that he makes no mention of Plato’s
allusions to games, and still more surprising that he says nothing about the
use of dice in any games. The only mention of dice in the whole passage
occurs in the quotation from Sophocles, and this does not necessarily mean
that dice were used in the five-lined game.

al
a
5
mo)
FIG. 13. FROM A TERRA-COTTA, ATHENS
(after Richter)

Very few game-materials have been found in Greece. A terra-cotta group

from Athens is pictured by W. Richter.! It shows a man watching two others
who have just finished a game on a board of 6 x 7 cells on which twelve men
are scattered. Falkener (p. 82) has a picture of an engraved gem at Naples
which shows two men playing a game on a board of 4x 5 cells, but on too
small a scale to permit of any exactness of detail. There are also diagrams of
ar8 boards, including that of Fig. 8B, on the steps of the temple at Eleusis? and on
the steps of the Parthenon and elsewhere in Athens. There is also a board of
the larger merels (Fig. 18F) on the steps of the Acropolis, Athens,? which may
troops. The omitted sentence is an obscure quotation from Cratinus, a writer of the fifth century
B.C. in which polis and kuon are mentioned. ' Spiele der Griechen und Rémer, 1887.
2 D. M. Liddell, Chessmen, 1938 (plate facing p. 115).
3 A. R. Goddard, ‘Nine Men’s Morris’, in Viking Club’s Saga-book, Coventry 1901. Goddard
suggests that this board may have been wrought by the same hands that cut the runic scroll on
the Lion of St. Mark on the top of the column at Venice when that animal stood on its original
site in the Piraeus.
2.5.1] GAMES IN THE ANCIENT WORLD 27
date from the classical period. Hyde (p. 205) says that the Greeks called merels
triodion, but his authority is late.
2.7.1. Petteia; polis, ‘city’ or poleis, ‘cities’. This was already played in
Greece by the fifth century B.c., and was probably still played in the second
century A.D., for Pollux’s description is so definite that it seems to be based
on actual knowledge of the game. It was a battle-game, played by two
persons on a board of some size with many pieces which possessed the move
of the rook in chess and captured by the interception method. Both Plato
and Pollux call the pieces “dogs’.' We shall see in Chapter 4 that games with
the same power of move and method of capture are still played in Egypt and
in Asia. These games suggest that the number of men was sufficient to fill
two rows at each end of the board.
I conclude that the petteia mentioned by Plato and the polis described by
Pollux were the same game, first because nothing that Pollux says is inconsis~
tent with what Plato says, and second because Plato (Rep. 422e), when com-
paring other states with his ideal state, refers to poleis as a feature of petteia.
Later commentators thought that the cities on the board were the cells on
which the men were placed. I think it more likely that the term ‘cities’ refers
to the massed body of men forming each of the contending armies. I find
support for my view in a passage in Aristotle’s Politics (12534) where he com-
pares the cityless man (apolis) to an azux. Azux was used later for the isolated
man in the game of tablé which is exposed to the risk of capture: the cityless
man is in danger because he is cut off from the support of his fellow-citizens.
2.7.2. Grammai, ‘lines’; diagrammismos. The only mention of this
game which the antiquaries found in the older Greek literature is a single
sentence in Philemon (mid-fourth to mid-third century B.c.): ‘he topes, he
plays diagrammismos, he dices (or plays kubeia)’. It is clear that Pollux knew
nothing about this game; he seems to have assumed that it belonged to the
same class as polis but played with fewer pieces. If this were so, Hesychius,
writing 400 years later, must be wrong when he says that it was played with
sixty men. I shall suggest below a possible explanation of this mistake.
Austin’s suggestion that diagrammismos was a race-game akin to the Roman
duodecim scripta is, 1 feel sure, right. The older Greek name, grammai, tallies
with the use of scripta in the Roman name, and to make the game a game of
skill like petteia makes nonsense of Philemon’s line. I am convinced that
grammai and duodecim scripta are the Greek and Roman names of the same
game, and I think that this gives a reasonable explanation of Hesychius’s
mistake. An explanation of the symbolism of alea, Gk. tablé (see 2.8.3), which
was popular from Persia to Spain two centuries after Hesychius,? was based

1 The pieces used in chess and tables were also named ‘dogs’ in Arabic by al-Ya‘qubi’s Ta’rikh,
and probably also by the Jews in Babylon (HC. 209, 446).
2 Th. Néldeke, who discusses the explanation in his ‘Persische Studien II’ in Sitzungsberichte
des K. Akademie der Wissenschaften in Wien., Phil.-Hist. Classe, vol. cxxvi (1892) Abhandlung xii,
p. 23, suggested that the explanation went back to a Neo-Platonic or Neo-Pythagorean source.
 28 GAMES IN THE ANCIENT WORLD lsh 2
on the calendar: the board has twelve divisions because the year has twelve
months; the number of pieces is thirty because the month has thirty days;
and the opposite faces of the die sum to seven because there are seven days
in the week. The board has, however, twenty-four, not twelve, points and
Hesychius may have reasoned that, as the number of points in the explanation
must be doubled, so must be the number of men.
2.7.3. Name unknown, but usually referred to as pente grammai, ‘five
lines’. The sole mention of this game is contained in the line from Sophocles
which Pollux has preserved, and this tells us nothing more than that it was
played on a board of five lines. It is Pollux who expounded the ‘five-lined
x board’ as meaning a square of 55 cells or points and dragged in the
proverb about the sacred line which has baffled scholars ever since.! Pollux
obviously knew nothing about this game.
Here, I think, the ancient Egyptian game-boards provide a better clue.
The only shape of board that satisfies the mention by Sophocles is the
pentagram, and the pentagram is one of the game-boards on the roofing
slabs of the temple at Kurna (see Fig. 7£). If the Greeks obtained their
games of petteia and kubeia from Egypt, it is possible that they obtained the
game on the pentagram in the same way. Moreover, a game on this board is
still played in Crete, and Miss L. S. Suther-
land, who saw it in 1938, gave me the
following description:
You have nine pebbles, and the aim is to get
them each on one of the ten spots marked. You
put your pebble on any unoccupied spot, saying
‘one’, and then move it through another, ‘two’,
whether this spot is occupied or not, to a third,
‘three’, which must be unoccupied when you
reach it; these three spots must be in a straight
FIG. 14. PENTALPHA, CRETE line. If you know the trick, you can do this one-
two-three trick for each of your nine pebbles and
find it a berth, and then you win your money. If you don’t know the trick, it’s
extremely hard to do it. The game is called pentalpha.
The spots are the intersections of the lines of the pentagram.
It is easy to see that the move must start from one of the angles of the
interior pentagon. A solution is:
1. B2—B1-A3; 2. A1—B3-Bz2; 3. As—B3-B4; 4. Bs—B4—A1; 5. A4—B1-Bs;
6. B3-B2-A4; 7. A2-B4—-B3; 8. B1-Bs5—Az2; 9. B1-B2-As.
This Cretan game is also played in northern India; in Sikkim and Assam
If this is true, the explanation was invented for grammai, and Hesychius could have been familiar
with it.
? The proverb is quoted by Alcaeus (c. 600 B.c.), Sophron (c. 430 B.c.), and Theocritus (3rd c.
B.C.), so the game from which it is drawn antedates both petteia and grammai. It has no parallel in
any known game and the descriptions by the antiquaries are contradictory. As Austin remarks,
‘the obscurity is impenetrable’ and modern attempts to reconstruct the game are unconvincing.
2.7.3] GAMES IN THE ANCIENT WORLD 29
where it is named Lam Turki,! and in the Karwi division, United Provinces,
as Kawwa dand.? The nine men are entered in the same way as in pentalpha,
and then the board is cleared as in solitaire (4.9.1). The pentagram is also used
in India, Central Provinces, for a hunt-game, kaooa (5.6.23).
A game on the pentagram is also played in Spain, name unrecorded, and
Spanish settlers in Mexico have taught it to the American Indians. The
Piman stock, Papago Indians, Pima City, Arizona, know it as ohohla, and the
Tanoan stock, Tews Indians, Santa Clara, New Mexico, call it akuyo,3 but
Culin gives no description of these games.
2.7.4. Boards, inscribed or scratched on steps and pavements, show that
the ancient Greeks also played games of the merels types.

8. ROME4
There are rather more references or allusions to board-games in the
classical Latin literature than have been found in Greek, and they are more
direct and less obscure. The board is generally called tabula, though abacus
and alveus are occasionally used, and Ovid (A.A. iii. 365) uses tabella for the
board on which the smaller merels is played. The piece is called calx by
Plautus, though later writers usually use its diminutive calculus generically,
and, more specifically, latro and latrunculus for the man in the Roman war-
game. The four-sided die is talus, ‘knuckle-bone’, and the cubic die is tessera,
and alea may have been used for the die or lot in the older literature, though
this is doubtful—in later Latin alea is the name of a particular game of chance
and it may have been a general term for any dice-game before it was restricted
to a single game of that kind. The game is Iudus, lusus, or focus (as in Ovid,
A.A. iii. 367). In later Latin, Iudus and Iusus must have passed out of use, for
the only term that passed into the Romanic languages was jocus (It. giuoco,
Sp. juego, Port. jogo, Prov. joc, and Fr. jeu). More specifically, we find ludus
(or Iusus) calculorum, ludus duodecim scriptorum, and Iudus latrunculorum, the
last played on the tabula latruncularia.s Plautus® already uses as a proverb or
metaphor the phrase ad incitas aliquem redigere, ‘to reduce somebody to
immobility’ which originally was used to describe a manceuvre in the game
duodecim scripta. There is a poem entitled Laus Pisonis, of the first century
A.D., in which Piso’s mastery at latrunculi is extolled, though neither the
terms Jatrunculus nor latro occur in it. At the end of the classical period, Bishop
Isidore (died A.D. 640) devoted a chapter of his Origines (xviii. 60 ff.) to the
game of alea or tabula, a shortened form of the duodecim scripta which had
taken the place of that game.
1 S. L. Hora, ‘Sedentary Games of India’, JASB., New Series, xxix (1933), p. 2.
2 E. de M. Humphries, ‘Notes on Pachesi’, JASB., New Series, ii (1906), p. 117.
3 Culin, Games of the North American Indians, Washington 1907, pp. 794, 798.
4 R.G. Austin has dealt with Roman games in his valuable article, ‘Roman Board Games’ in
Greece and Rome, iv (1934-5), 24-34, 76-82. See also S. G. Owen, Ovid, Tristia IT (Oxford 1924),
pp. 252 ff., and H. Lamer in Pauly-Wissowa, s.v. Lusoria tabula.
5 Seneca, Ep. cxvii. 30. ® Poen. 907.
30 GAMES IN THE ANCIENT WORLD [CH. 2
The literary evidence is supported and clarified by the survival of many
game-boards, which have been found in Rome itself and in other parts of
the Roman Empire. Many of these have been found on Roman sites in
Britain. These are of three types:
2.8.1. Name unrecorded; the smaller merels. Boards have been found
at Timgad in North Africa.! Ovid? mentions this game:
Parva tabella capit ternos utrinque lapillos,
In qua vicisse est continuasse suos.

The game is described in Chapter 3 (3.3.1-19).

2.8.2. Ludus duodecim scriptorum; duodecim scripta, ‘twelve lines’.
Cf. the Greek grammai (2.7.2). A board of 3 x 12 cells (technically known as
‘points’). The points are generally marked by circles, squares, vertical bars,
letters, monograms, leaves, crosses, crescents, and most frequently by letters,
the whole series of which form a sentence. The sentences fall into three types:
(a) maxims for players, (b) jeers at ‘rabbits’, and (c) references to the circus,

VENARI LAVARI
IV DPE Ee SReISDEE RE
O.C CabrS ale meV leVabpicce
FIG. If. BOARD FOR DUODECIM SCRIPTA (from Timgad)

eating, and other pleasures.3 N. I. Lanciani, writing in 1892, said that more
than a hundred boards had been found in Rome alone during his lifetime.4
A board, dating probably from the first half of the second century A.D., was
found at Holt, Denbighshire, and is now in the National Museum of Wales.
It is made of buff ware; the twelve points on the two outer rows are marked
by heart-shaped ivy leaves and those on the middle row by monograms. The
two ends of the bar are marked by semicircles and its centre by a circle with
a six-armed rosette.5 A finer board with the same markings was found in
1931-2 during the excavation of the tombs of the Blemye kings and nobles
at Qustul, a little north of the Second Cataract of the Nile. This board,
77's by 37 cm., is made of a single piece of wood with a framed border and
corner brackets of silver, and the points are marked by small squares of ivory
fretwork inlay. With the board were found thirty pieces, half ivory and half
ebony, five ivory cubic dice, and the fragments of a pyrgus for throwing the
dice and of the leather bag in which they were kept. The tomb is later than
A.D. 291 and not later than the sixth century A.p.°
x, See Fi
™ Austin, a, 79. @) A.A. iii. 365-6; cf. Trist. ii. 481-2.
3 Austin, a, 32, n. I. 4 Culin, d, 832.
5 W. F. Grimes, Y Cymmrodor, xli (1930), 131 and R. G. Austin, Archaeologia Cambrensis,
Dec. 1938.
© W.B. Emery, Nubian Treasure, London 1948, p- 46 and pl. 32.
 2.8.2] GAMES IN THE ANCIENT WORLD 31
More illuminating, since it obviously indicates the order in which the men
are entered, moved, and borne in the course of play, is a board from Ostia,
Italy?
CEeCCCeEe BBBBBB
AAAAAA AAAAAA
DDD DDiDS SE-ELE Eee
FIG. 16. BOARD FOR DUODECIM SCRIPTA (from Ostia)

I have already expressed my conviction that duodecim scripta is the same

game as the Greek kubeia or grammai. Both are race-games of the backgam-
mon type, played between two persons on a board of 3x12 points, each
player having fifteen men whose moves are governed by the throws of three
cubic dice. The Ostia board shows that the men were entered in table A,
moved through the tables lettered B, C, and D, in that order, and were
borne from table B when the player had brought all his fifteen men to that m
table. The first player to bear all his men won. Each player in turn threw the
three dice, the throws could be used singly or combined, but could not be
divided and a throw of doublets or triplets gave no second throw. Blots
(single men on a point) were called vagi and could be taken and then had to
be re-entered before any other man could be moved. A player could pile
any number of his men on any point; piled men were called ordinarii and
were immune from capture so long as they remained piled. Men that could
not be moved by any throw of the dice were called inciti.2 The rule permitting
piling involved the use of disks for the men instead of the pawn-shaped men
used in the Egyptian games; this would seem to have been adopted by the
Greeks.
2.8.3. Alea or tabula. A Roman modification of duodecim scripta in which
the third row of points on the board is omitted, thus shortening the track.
Otherwise the method of play remained unaltered. We do not know when
this change was made; it is possible that some of the earlier mentions of alea
may have related to the reformed game, but it came to the front in the first
century A.D. when it was a favourite game of the Emperor Claudius (41-54).
Suetonius says that the Emperor had written a book on the game and
had a board (alveus) so fixed to his carriage that he could play alea on his
journeys. The game became so popular that it obtained a second name,
tabula, a transference of meaning from the name of the board which has
many parallels. Its use became so general that when alea spread to Byzantium,
it passed into Greek in the form tablé. In the Western Empire and in medieval
Latin, alea survived in legal and ecclesiastical documents and in works of
scholarship as the proper name of the game, but, almost invariably, it is
explained or glossed by the name in common use, alea, id est tabula. Thus in
Oa CIL., xiv. 5317; cf. Austin, a, 33. ? Isidore, loc. cit. 3 Claudius, 33.

Yr
32 GAMES IN THE ANCIENT WORLD (CH:2
Canon 79 of the Synod of Elvira, Spain (c. A.D. 305), we read: ‘Si quis fidelis
aleam, i.e. tabulam, luserit numis, placuit eum abstineri.’ Here it is the playing
of alea for money that is condemned, not the game itself. But it was not long
before stricter control, first of the clergy and later of the laity, became
necessary.
In Justinian’s Code (729-32) the clergy are forbidden ad tabulas Iudere (1. vi. 17: in
the Novellae, tablizare) and Bishop Isidore ends his valuable but greatly misunderstood
account of ‘alea, i.e. ludus tabulae’ thus:
68. De interdictione aleae. Ab hac arte fraus et mendacium et perjurium nunquam
abest, postremo et odium et damna rerum, unde et aliquando propter haec scelera
interdicta legibus fuit.

Four epigrams of Agathias of Myrine, a contemporary of Justinian, in the

Palatine Anthology (ix. 482, 767-9) have the game of tablé for their subject.
The last three deal with the effect of the game on players and on the domina-
tion of chance which makes it a picture of man’s life, but the first describes
a position, which occurred in a game played by the Emperor Zeno (475-81),
in such detail that it invites reconstruction. This was attempted by Hyde,
Becq de Fouquiéres, and Henry Jackson with varying results. The whole
matter was re-investigated by Austin,! who has established that Becq de
Fouquiéres’s solution is the only one that is in complete harmony with the
epigram.

Zee
ex
e
@

12)
°
ce)
12)
°
ie)
°

if
FIG. 17. ZENO (WHITE) TO PLAY 2, $, 6

In the position diagrammed, Zeno, playing with the white men, threw 2,
5, 6. He was therefore compelled to unpile his men on three of the points
held by his ordinarii, since his ordinarii on f and his vagus on j are all inciti for
this throw, with the result that the throw left him with no fewer than eight
vagi and a ruined position.
* “Zeno’s Game of raBAn’, JHS., liv (1934), 202-5.
2.8.3] GAMES IN THE ANCIENT WORLD 33
Zeno’s game shows that white entered in af, moved in direction amnz and
were borne in fz. It tells us nothing about black, but it seems reasonable to
assume that both players entered in the same table, traversed the board in
the same direction, and bore their men from the same table, as shown in the
Ostia board for duodecim scripta, though this is not certain, for other arrange-
ments and directions of move were used later.!
2.8.4. Ludus latrunculorum, latrunculi, ‘soldiers’. Played by two per-
sons on a latticed board marked with ‘lines and spaces’,? but the exact size is
nowhere specified. Many boards which satisfy Varro’s description have been
found on Roman sites in Britain, particularly along the Roman Wall. Three
boards from Cilurnum with 7 8, 8 x 8, and 9x 10 cells, respectively, are in
the Chesters Museum with a number of button-shaped men. A board of
7X 8 cells and fragments of two other boards and a few game-counters were
found at Corbridge in 1911, and boards have also been found in some of the
four-mile castles and turrets on the Wall. At Richborough, a board of 8 x 8
cells, each cell having its diagonals added, and a fragment of a board of
10X13 cells have been found with many flat disks; one group of eighteen
had eleven marked with concentric circles and seven plain. At Chedworth,
Gloucestershire, a board of 8 x 8 cells was found. From these boards it would
appear that the exact dimensions did not matter, but that 8 x 8 cells was the
norm.
No reference says anything about the number of men in the opposing
armies, but the use of the term mandra (used by Juvenal for a drove of cattle)
implies that the number must have been considerable and probably sufficient
to fill two or possibly three rows on either side. There was no differentiation
of piece; all the men had the same powers of move and capture. Captures
were made by the interception method, and moves in the orthogonal direc-
tions, forwards or backwards, to right or left. The method of capture entails
the use of the rook’s move in chess; that a man could move farther than to an
adjoining cell seems implied in the Laus Pisonis.3
Latrunculi was carried by Roman legionaries to the frontiers of the Empire,
as may be seen from the number of materials for the game that have been
found in Britain. The last references to it as still played in the interior of the
Empire that I have found belong to the fourth and fifth centuries a.p.
Vopiscus (c. 305), in Procul. 13. 2, says that the winner of a game of latrunculi
’ The later history of alea in western Europe will be given in Chapter 6. In medieval Greek
tablé was known as taula and this (Ar. tawula) has been adopted by the Levantine Muslims. Trade
also carried tablé to Russia and was followed by canonical intolerance. The early Russian Svodni
Kormch extended the ban on the game to the laity, laying down that ‘no clergy nor layman shall
play at zerniyu (hazard), shakhmate (chess), or tablei (alea). In 1561 the Tsar Ivan IV made fablei
illegal under the civil code, and the game passed out of use until games on the backgammon board
were reintroduced from western Europe in the reigns of Peter the Great and his successors.
(E. V. Savenkov, K. Voprusu op Evolutsie Shakhmatnoi Igry, Moscow, 1905, pp. 74-77, &c.)
“Varro, axa 22s
3 This paragraph summarizes the conclusions reached by Austin in his article on ‘Roman Board
Games’, 25-30.
6356 D
34 GAMES IN THE ANCIENT WORLD [chs
was hailed as ‘Imperator’. Macrobius (c. 400), in Sat. i. 5, rebukes those who
played at tabula and latrunculi. In Rome and Italy the game perished with the
fall of the Roman Empire: the view, often expressed, that Jatrunculi was the
parent of draughts is unsupported by any evidence and is, I believe, untenable.
There is nothing in all that we know about Jatrunculi that is inconsistent
with anything that we know about petteia. So far as the evidence goes,
latrunculi is petteia. We know too little about the Greek grammai to say the
same thing about the two race-games, but the relationship between latrunculi
and petteia does suggest that duodecim scripta and grammai were also the same
game. We have seen that the duodecim scripta board is identical with the Egyp-
tian board from Ak-hor, and we have Plato’s definite statement that the
Greek games of petteia and kubeia came to Greece from Egypt. The Egyptian
ancestry of the games played in ancient Greece and Rome seems established.
Before I pass on to the description of the games of the different classes
enumerated in Chapter 1, it will be convenient to deal in brief with the
oldest games of other parts of the world of which we know little more than
their names.
9g. CELTIC GAMES
Board-games are frequently mentioned in the earlier Welsh and Irish
literature, but few details are given and no game-materials of the older
games are known. Dice or lots are never mentioned in connexion with these
games, so we may infer that they were games of skill alone. The game that
is most frequently mentioned is
2.9.1. Gwyddbwyll, Wales; Fidchell, Ireland. These two names are
linguistically identical: Welsh gwydd = Irish fid, both meaning ‘wood’.
Gwyddbwyll is mentioned several times in the Mabinogion' and fidchell in the
Irish romances. In both countries the game seems to have been supplanted
about A.D. 1000 by the Norse game, hnefatafl (4.1.13), which was known
in Welsh as tawllbwrdd, an Irish board for which was found in 1932. Most of
the mentions in the Mabinogion simply record that two people are playing
gwyddbwyll on a silver board with men (gwerin, werin) of gold. But in the
story of Peredur, he meets with a magic board on which the pieces are
playing by themselves. Lady C. Guest, in her translation of the Mabinogion,?
quotes from another work: “The chessboard (gwyddbwyll) of Gwenddolen;
when the men were placed upon it, they would play of themselves. The
board was of gold and the men (gwyr) of silver.’
Fidchell is included in Cormac’s Glossary (c. 900, but with later additions),
and the entry, after an attempt to explain the word, gives the following
account: ‘In the first place the fidchell is four-cornered, its squares are right-
angled and black and white men are on it, and moreover it is different people
that in turn win the game.’ A much later passage in the Book of Lismore
’ Red Book of Hergest, 1887, i. 84, 153, 158, 220, 221, 240. 2 1849, 1. 383.
3 Ed. W. Stokes, 1868, p. 75.
2.9.1] GAMES IN THE ANCIENT WORLD 35
(1sth cent.) describes the gift of a fidchell to Pope Boniface in the seventh
century, and says that it had nine lines and half of the pieces were men and
half women, but by the time this work was written, fidchell had been long
obsolete and had been generally equated with chess.
Gwyddbwyll (fidchell) was evidently a battle-game, played by two persons
each with undifferentiated men. I suggest that it was the same game as
latrunculi which we have seen had reached Britain during the Roman occupa-
tion and may easily have been learnt by the native population and spread to
Ireland.
2.9.2. Irish texts also mention three other board-games which were
obviously games of skill though no details are given. These games were
cennchaem (or cendchaem) Conchobair, ‘Conchobar’s Fairhead’, bran-
dub (brandubh or brannaib), and buanfach.

10. INDIA
Board-games seem to have been late in appearance in India. H. Liiders
says that there is no record of their existence in the Vedic period, when
gambling with lots and chariot-racing are the only forms of recreation men-
tioned. The earliest mention of board-games appears to be that in the
Brahma-jala sutta! which purports to record the actual words of Gotama,
dating back to the fifth century B.c. In this work Gotama enumerates the
trifles which occupy the thoughts of the unconverted man, and among them
is playing ashtapada and dasapada. Ashtapada is also mentioned in the Hari-
vamsa? where a game between Rukmin and Balarama ends in a quarrel in
which the board is used as a weapon.3 The Mahabhashya (end of the second
century B.C.) defines ashtapada as ‘a board in which each line has eight
squares’, i.e. is the 8 x 8-celled board, and dasapada is the 10x 10 board. Both
were race-games and the ashtapada game is still played in India (see Chapter 6)
but on boards of smaller dimensions. In the early centuries A.D. we meet with
references to another race-game, chatush-pada, the game now known as
chaupur, and a carving in the Brahmanical cave-temples at Elura (A.D. $79-
725) shows this game. The outstanding event in the history of Indian board-
games is the invention of chess in the second half of the sixth century .p.,
after the end of the Hun domination, 458-540, which had shaken Indian
society to the foundations and severed all traditions.‘

Il. CHINA
If any reliance could be placed on the articles on board-games in Chinese
and Japanese encyclopedias, we should have to admit that board-games were
already played in China in the third millennium B.c., since all agree in saying
that wei-k’i (4.7.1) was invented between 2357 and 2254 B.c. But Chinese
1 Ed. Rhys Davids in the series Sacred Books of the Buddhists, 1899, 1. .
2-H. Liiders, 65. 3 See HC. 35, where other references to this game are given.
4 V. A. Smith, Early Hist. India, 3rd ed., 1914, p. 410.
 36 GAMES IN THE ANCIENT WORLD [CH..2
historians have always tended to exaggerate the age of their inventions, and
in particular the age of their games. Modern scholarship holds that the only
Chinese board-games before the Christian era were simple games of the
merels type, i.e. games of alinement. The yih mentioned by Confucius (551-
479 B.C.) and Mencius (372-289 B.C.) was the smaller merels (3.3.1-19). Yih,
like the term k’i which has superseded it, means a board-game in general.
During the first millennium a.p. Indian race-games began to reach China.
Karl Himly, who did much research in the history of Chinese games, quotes!
from the Hun Tsun Sii, a work of the Sung period, 960-1279, a passage
about the game t’shu-p’u which says that it was invented in western India and
spread to China in the time of the Wei Dynasty (a.D. 220-265). T’shu-p’u is
in fact the Chinese adaptation of the Skr. chatush-pada (mod. chaupur). The
passage goes on to say that the game had had four other names in succession
in China, wu-sho, ‘spear-seizing’, thshan-han, ‘long row’, po-lo-sai-hi, and
shwan-liu, “double sixes’. The last of these names, however, belongs to a
different game from chaupur; it is the Persian and Arabic nard, which is
affiliated to the Greek tablé (2.8.3) and, according to Persian tradition, was
introduced into north-west India towards the end of the sixth century A.D.
Chess was introduced from India c. 700, by the ancient trade route from
Kashmir. The oldest and best of the native Chinese games, wei-k’i, is not
older than A.D. 1000.
? ZDMG., xxvii, 1873, p. 479; cf. HC. 120.

Additional Note on Ur (pp. 19 ff.)

While this chapter was passing through the press, I learnt from Sir L. Woolley,
that from time to time during his explorations at Ur, bricks, of face 12 inches by
9, and ranging in date from 2000 to 750 B.c., were found. In one face of these
bricks four rows of holes had been roughly ground. So far as he remembers, the
two inner rows each contained eight holes and the two outer rows contained in
their middle three holes. A similar brick had been found by him at Carchemish
on the middle Euphrates. It is difficult to see any purpose which these bricks can
have served other than as a game-board.
 S.
GAMES OF ALINEMENT AND
CONFIGURATION
Gags of alinement are played in Europe and Asia on lined boards. Two per-
sons play, and begin with their men in hand, and have to enter them on the
points of the board, one at a time in alternate turns of play, with the aim of
alining three men on the three points of a marked line of the board. This
arrangement is known in Britain as a row.’ In Europe during the Middle Ages
it was known in the Romanic languages as a line or table; ML. linea or tabula;
MF. ligne, table, or taule; It. filo or filetto; and in the Germanic languages as
a mill; Ger. miihle; Dan. mille; Ic. mylna; Sw. qvarn. In some parts of Africa,
a game of alinement is played on the cells of a latticed board.
The Japanese also play a game of alinement on the cells of the wei-k’i (Jap.
i-go) board, in which the aim is to aline five men, and this game was intro-
duced into Europe at the end of the nineteenth century. I shall classify the
games by the number of men required to form a row.

THREE-IN-A-ROW GAMES
The boards on which these games are played are shown in Fig. 18. The
games are very old. We have seen in Chapter 2 that boards have been found
at Kurna in Egypt, Crete, Greece, and Rome, and that the games were
played already in China c. 500 B.c. Boards have also been found incised on
rock near ancient sites in Ceylon. Today they are widely spread over the
world, and may be even more widely spread, since their simplicity may have
resulted in their escaping notice. One or other of the games has been recorded
for most European countries for all parts of Asia except Siberia, Burma,
Siam, Annam, and Indonesia, for East Africa as far south as Somaliland, in
the Sudan, in parts of West Africa, and in Madagascar. Some American
Indian tribes have learnt them from Spanish settlers.
The three-in-a-row games were known in Arabia as irq, qirqa, a term
which is already used in the Kitab al-aghdni (ante A.D. 967),? where it is
mentioned that an inhabitant of Mecca kept sets of chess, nard, and girq for
his guests’ entertainment. Qirg is not derivable from any Arabic root and
must have been borrowed from some other people, presumably with the
game itself. The word is included in many Arabic dictionaries, and generally
(as in the Qam#s, ante 1414), accompanied by a drawing of board F or G
(Fig. 18). The Moors carried the word to Spain and it passed into Castilian
' Dialectally in Warwickshire, click-clack. ? Ed. Bulag, iv. 511.
38 GAMES OF ALINEMENT [CH. 3
Spanish as alquerque, which is used in Alf., not only for the three-in-a-row
games, but also for two other games, one a war- and the other a hunt-game,
which were played on lined boards, so girq may have originally meant simply
a lined board. This is supported by the use of alquerque in Spanish as the
name of that part of the oil-mill in which the bruised olives are laid out,
which has channels or grooves to catch the oil. Qirq is now obsolete in
Arabic and has been replaced by the word dris (often confused with Edris, the
name of the patriarch Enoch).
oo

op Ms

Ouse)
A B

E
FIG. 18. BOARDS FOR THREE-IN-A-ROW GAMES

In the Romanic and English languages the generic names for these games
are derived from the Low Latin merellus, a ‘token’, ‘coin’ or ‘counter’.! The
English word, merels, has undergone many dialectal perversions of which
one, morris, has largely taken its place. I have noted the following:

madell (Wilts.), marells (Dorset), marl (Wilts.), marlin (Hyde, 204), marls, marnull
(Dorset), marriage (Derby), marrel, marril, maulty (Forest of Dean), medal (Wilts.),
merls, merrils (Wilts., Yorks.), merryal (Lincs.), merryholes (Lincs., Northants), meg
merrylegs (Lincs.), merrymen, merrypeg (Oxon.), peg meryll (Northants), miracle
(Oxon.), miracles (Ches.), miraele (Oxon.), morals (Hyde, 204), moris (Hyde, 204),
morrell, morris (Glos., Middx., Norfolk, Northants, War., Worcs., Yorks.), morrit,
multi, murrells, mutti (Forest of Dean).

* The plural merelli was also used for other games with undifferentiated men, e.g. in English
for fox and geese, in French for the war-game alquerque which La Vieille (see p. 3, note 1) calls
‘le gieu des merelles qui se fait par douze merelles’, and for hop-scotch (mod. Fr. mérelle).
cig AND CONFIGURATION 39
Another European name for these games is mill, used in the Germanic
languages and thence adopted in many Slavonic and Eastern European
languages: Ger. miihle, miihlen, miilchen; Du. molenspel; Dan. mélle; Ic. mylna;
Cz. mlyn; Russ. melnitsa, melnichny; Hung. malomjatek, malmosdi; Sw. qvarn.
Brand! says that the games are called Shepherd’s mill in some parts of Eng-
land, and modern French and Italian dictionaries give Fr. jeu de moulin and
It. molino, mulinello, as occasionally used.
In Catalan marro is used, and in parts of Spain, pedreria which has been
borrowed by various American Indian tribes. The Yorubas of the Gold Coast
and Nigeria call the games akidada.
I. NINE HOLES. Played on boards A, B, and C (Fig. 18). Two players,
each with three men in hand which are entered, one at a time, in alternate
turns of play. The aim is to make a row orthogonally and the first to make
his row wins. When all the men are entered, the game proceeds by alternate
moves in which a man can be transferred to any vacant point or cell.
3.1.1. Britain: Nine holes. Played on boards A and B. One or other
board may be seen in the cloisters of practically every English cathedral of
monastic origin. At Westminster Abbey, the holes are exceptionally large
and deep, and show every sign of much use. The game is not to be confused
with two other games bearing the same name: one, a game of marbles,” the
other a ball game resembling rounders.3 References are usually too vague to
show which game is intended, but the record of a Manx ecclesiastical court
in 1699, punishing two men for ‘Makeing Nine Holes with their Knives after
Evening Prayers’ clearly refers to the merels game.
3.1.2. Holland: Driesticken (Hyde, 204).
3.1.3. Germany: Kleine miihlenspiel (Fiske, 129, citing Archiv der Spiele,
1819-21, ii. 21-27); Nulochen (Hyde, 204); Trip trap trul, in Lower
Saxony (K. Niebuhr’s Reisebeschreibung, Kopenhagen, 1774, i. 171).
3.1.4. France: Les Pendus (M. Kraitchik, Mathematical Recreations, Lon-
don 1948, 291): Played on board D. Rows may also be made diagonally.
3.1.5. Arabs: Dris ath-thalatha (Carsten). Board C.
3.1.6. India, Punjab: Tre-guti (H. C. D. Gupta, “Games from the Punjab,
Central Provinces, and Orissa’, JASB., New Series, xxii (1926), 143).
Board B.
3-1.7- Ceylon: Nerenchi (Parker, 579).
3-1.8. Japan: San-noku-narabe (Prof. Tsuboi). Board B. Played by
children and by adults in remote villages. The only game played by the
Ainus, who name it chikkiri.
3.1.9. Ancient Egypt: name unknown. Board B at Kurna; board C of the
Roman period in the British Museum, no. 14315.

! Observations on popular Antiquities, 1777.

2 J. Strutt, Sports and Pastimes of the People ofEngland, edited W. Hone, 1838 (ed. London 1876),
op. 368, 491.
3 Country Life, 26 Oct. 1946. 4 W. Cubbon, Times, 21 Novy. 1929.
40 GAMES OF ALINEMENT [CHa
3-1.10. Somaliland, Issaq tribe. Jara-badakh (Marin, a, 505). Board B.
Played by children.
3.1.11. Gold Coast, Akuna tribe; Nigeria, Yoruba tribe: Akidada (K. C.
Murray, who saw it at Nopa). Board B drawn on the ground. ‘Sticks’ made
from the fibres of palm leaves are used as men, each player sloping his men
towards himself to distinguish the sides.
2. NOUGHTS AND CROSSES. Played on board C, but the outer boundary
of the board is usually omitted and the players, who make the board afresh
for each game, form it by drawing two parallel lines vertically and two
horizontally across them. A children’s game, played in Britain, some parts of
Europe, and by European settlers in America. The board is drawn on paper,
slates, or on the ground. Two players. The game is one of entry only and the
men have no power of movement; their positions on the board are marked
by one player with an ‘O’ and by the other with an ‘X’ in Britain and a ‘1’ in
Sweden. The aim is to make a row in any direction. When a player succeeds
in making a row, he draws a line through the three men, exclaiming in
Britain “Tit tat toe, Here I go, Three jolly butcher boys all in a row’, and in
Sweden “Tripp trapp trull, Min qvarn ar full’. In England, drawn games are
scored to Old Nick or Old Tom. Nothing is known as to the antiquity of
this game. It was more frequently played in the days when slates were used
in schools than at the present day.
A. C. White! has shown that the first player has only three different moves
(corner, middle of a side, and centre) and that there are only twelve positions
after the second player has moved. In six of these the first player should win
and in six should draw. The second player can never win unless the first
player makes a mistake.
3.2.1. Britain: Noughts (Oughts) and crosses. Also known as Kit cat
cannio in Suffolk (A. Gomme, Games traditional, London 1894, i. 311), Tip tap
toe in Lincoln, Tick tack toe in Warwick and Stafford, Tit tat toe in Lincoln,
Berkshire, and in the United States (Fiske, 127).
3.2.2. Holland: Tik tak tol (Fiske, 122 n.).
3.2.3. Sweden: Tripp trapp trull (Fiske, 137-8).
3. THE SMALLER MERELS (Three men’s morris). Played on board D.
Two players, each with three men in hand. The men are entered, one at a
time, in alternate turns of play, each player endeavouring to form a row
along one of the eight marked lines of the board. When all the men have been
entered without a row being formed, the game proceeds by alternate moves in
which aman can be moved onestep along any line through the point on which it
stands to a neighbouring empty point. The player who first makes a row, wins.
The game is fully described in Alf. 934, where it is shown that the first
player has a forced win if he enters his first man on the central point b2. His
opponent must enter his first man either on a corner point or on the middle
point of an edge, and the game proceeds as follows:
“Tit tat toe’, Brit. Chess Mag., 1919, p. 217.
3.3.1] AND CONFIGURATION 41
1 ob2, if 293, ba ab3; 3.cesj,42%45 br-c1, any; 5: b2-<2
if b3; 2. al, ¢3; 3. a3, a2; 4. di—b1, any; 5. br-cr
where the moves in italics are all forced. Moulidars and other French game-
manuals accordingly prohibit the first player to enter his first man on bz.
This rule, however, is not adopted outside France.
3-3-1. Greece and Rome: name unknown. See Chapter 2.
3-3-2. Spain: Alquerque de tres (Alf. 932); Tres en raya; Castro; collo-
quially, Pedreria.
3-3-3. Catalonia: Marro (Brunet y Bellet, El Ajedrez, Barcelona, 1891,
204).
3-3-4. Italy: Smerelli; mulino; semplice mulinello; filo; filetto;
tavoletta; riga di tre (Carrera, I] Gioco de gli Scacchi, Militello 35). Now
only played by children.
3-3-5. France: Mérelles; marelles (Fiske, 111-17); carré chinois (Léon
de la Borde, Notices des émaux du Musée du Louvre, 1853, ii. 381).
3-3-6. Britain: Three men’s morris. The ordinary name, but most of
the perversions given above are used dialectally, and to distinguish this game
from other merels games, the generic name is often preceded by three men,
three pin, three penny, or three peg. Hyde (p. 211) adds for Cumberland
copped crown, and Gomme (i. 80) for Galloway, corsi crown.
The game only reached Britain after the Norman Conquest, but was well
established by 1300. Board D was often drawn on the cloister seats of cathe-
drals, as at Norwich, Canterbury, Gloucester, Salisbury, and Westminster
Abbey. I have found it scratched on the desks of old choir stalls at Soham
(Cambs.) and Ludlow (Salop). During the restoration of Hargrave Church
(Northants) in 1869, it was found on a stone built into the wall of the north
aisle; it may be seen ona stone of the chancel arch at Singleton (Sussex), and
on a tombstone at Arbory in the Isle of Man. It was much played by shep-
herds and boys when watching sheep, but is less frequently played now in
England.
3.3-7- Ireland: Cashlan gherra (Hyde, 211). W. G. Wood-Martin (Lake
Dwellings in Ireland, p. 130) quotes T. Crofton as saying that three men’s
morris was widely played in Ireland in the early nineteenth century.
3-3-8. Arabs: Dris (formerly qirq), ath-thalatha (Dozy, Supplément aux
dict. arabes). Seen and described by K. Niebuhr, Reize naar Arabie, Christiania
1776, 1. 168.
3.3.9. Persia: Hujura (Hyde, 211).
3-3-10. India; Lower Bengal: Tant fant (J. M. Datta, 167). The men are
arranged (one player, a1, bi, c1; the other a3, b3, c3) and not entered on the
board. The aim is to make a row on any other line. Cf. 3.3.18 below.
3-3-11. India, Vikrampur, E. Bengal: Tin-guti pait pait (tin = ‘three’.
E. de M. Humphries, 117).
3-3-12. Ceylon: Nerenchi (Parker, 507). A board cut on a rock near the
Lankara dagaba. Now obsolete.
42 GAMES OF ALINEMENT [CH. 3
3-3-13. China: Luk tsut k’i, ‘six men game’ (Hyde, 211, who trans-
literates the name /o che). Played already in the time of Confucius under the
name yih. Named kiu kung under the Swei-shu dynasty (A.D. 581-617)
when twenty-one books were written on the game.! Kiu kung, ‘nine castle’,
still survives as the name of that part of the Chinese chessboard which carries
the same design, and to which the general and counsellors are confined. The
three-in-a-row game is said to be confined now to southern China.
3-3-14. Philippines: Tapatan (Culin, f, 648). Widely played.
3-3-5. Ancient Egypt: name unknown. Boards at Kurna (see Fig. 7).
3-3-16. Abyssinia: Santaraj, ‘chess’; ‘Pilate’s game’ at Choa (Griaule,
Jeux et divertissements abyssins, Paris 1935, p. 176).
3-3-17- Somaliland: Carabawg (Griaule, a, 176).
3-3-18. Nigeria, Warri Province, Urhobo tribe: Epelle (K. C. Murray,
who saw two boys playing it at Owu). Played in the same way as 3.3.10.
3-3-19. Madagascar: Fanorona telo (Mrs. Danielli). Played by Imerina
children.
3.3.20. American Indians; Tanoan stock, Tewa Indians, Santa Clara, New
Mexico. Pitarilla or Bidaria; Tigua Indians, Isleta, New Mexico: Picaria,
where the middle points of adjacent sides are joined diagonally. All these
names are perversions of the Spanish pedreria (Culin, g, 797-8).
4. FIVE (OR SIX) MEN’S MORRIS. Played on board E. Two players, each
with five (or six) men in hand. The men are entered, one at a time, in alter-
nate turns of play, each player endeavouring to form a row along one of the
marked lines of the board. Every time a player succeeds in making a row, he
takes one of his opponent’s men, choosing any man that he likes, and removes
it from the board. When all the men are entered, the game proceeds by
alternate moves in which a man can be moved one step along any line
through the point on which it stands to a neighbouring empty point. The
player whose men are reduced to two, loses.
The game was popular in Italy, France, and England in the Middle Ages,
but dropped out of use before 1600. Today it only survives in West Africa,
where, in addition to the normal method of capture, a player who secures
two men on any line connecting corresponding points on both squares takes
one of his opponent’s men.
3-4-1. Italy: Smerelli. The only evidence for this game is a diagram which
a sixteenth-century player added on a blank page (f. 2182) of his copy of the
CB. work. This manuscript was formerly in v. d. Lasa’s library, Storchsnest,
Prussia, but was sold in 1935, and its present ownership is unknown.
3.4.2. France: Mérelles a six tables. Mentioned in a document of 1412 in
the Paris archives, quoted by Ducange (iv. 157): ‘Jean Aysmes qui avoit joué
aux merelles 4 six tables, appelé le jeu Saint Marry’ (which Ducange latinizes
as Sancti Mederici).
' K. Himly, ‘Anmerkungen in Beziehung auf das Schach und andere Brettspiele’, ZDMG.,
XXvii (1873), 475, gives the titles of these works.
3-4.3] AND CONFIGURATION 43
3-4.3- England: Fivepenny morris. Mentioned in Cotgrave: ‘Merelles.
Le jeu des merelles. The boyish game called Merills, or fiue-pennie Morris:
played here most commonly with stones, but in France with pawnes, or men
made of purpose, and tearmed Merelles.’ Hyde (p. 204) has the names ‘five
penny moris’ and “five pin moris’ without describing the game. A passage
in Apollo Shroving (1626) quoted in Gomme (i. 12), s.v. ‘“Fippeny Morrell’,
confuses this game with the smaller merels: “Twice three stones, set in a
crossed square, where he wins the game that can set his three in a row, and
that is fippeny morrell I trow.’
3-4-4. Gold Coast: Achi or Ati; Nigeria (Yoruba tribe): Akidada (K. C.
Murray, who saw it played at Nopa in 1928 and 1929). Two players, on the
first occasion Yorubas, and on the second from the Gold Coast, were playing
on a board traced on the sand, each having six ‘sticks’ made from the fibres
of palm leaves, one side green, the other brown. On the second occasion he
learnt the rules of the game as given above.
A different game, played on a board consisting of three concentric triangles,
the corresponding angles of which were joined by lines, was described by
Mr. G. E. Dartnell.!
3.4.5. England, Wiltshire: Sixpenny madell. Two players, each with six
men in hand. Since the board has only six points, the mid-points of the sides
of the triangles must be regarded as points on which men can be placed.”
5. THE LARGER MERELS (Nine, Eleven or Twelve men’s morris).
Played on boards F and G. Two players, each with nine men on board F and
eleven or twelve on board G. The men are entered, one at a time, in alternate
turns of play, each player endeavouring to form a row along one of the
vertical or horizontal lines of the board and, in some games on board G, on
the slant lines, or to confine his opponent’s men so that they are unable to
move. Every time that a player forms a row, he removes one of his oppo-
nent’s men from the board: in some games this may not be taken from a row.
When all the men are entered, men can be moved one step along a line
through the point on which they are stationed to a neighbouring empty
point. The player wins who succeeds in blocking all his opponent’s men or
who reduces their number so that they can no longer form a row. These are
the standard rules; variations will be given when I deal with the games
individually.
The game was originally played on board F (MF. merellier) and this is still
the more usual board. At a later date, but not later than 14003 the diagonal
lines were added and, at first, rows could not be made on a slant line. This
change provided no more points, but led eventually to an increase in the
number of men in each player’s possession from nine to eleven or twelve.
l Notes and Queries, 8th ser., xii. 333.
2 The analogy of other merels games suggests that the medians of corresponding sides of the
triangles and not the corresponding angles should be joined, and experiment shows that this
would give a much better game.
3 Board G is already diagrammed in al-Firiizabadi’s Qamiis (ante 1414).
44 GAMES OF ALINEMENT [CH. 3
Board F is very old and we have seen in Chapter 2 that it already occurs on
the roofing slabs at Kurna, and may have existed in Greece. In Ceylon
(Parker, pp. 507 ff.), two boards are cut on the great flight of steps up the hill
at Mihimtali, and others on a rock near the Lankarama dagaba; these are not
later than the first century A.D. A board inscribed on a flat stone from the
burial site of Cr Bri Chualann, co. Wicklow, Ireland, belongs to the Stone
and Bronze Age.! A fragment of a wooden board, on one side hnefatafl
(4.1.13) and on the other board F, was found in Scandinavia in the Gokstad
ship of the Viking Age. Hyde (p. 203) says that the game (mereles) is men-
tioned in the Talmud.
Board F was often scratched or incised on stone in Britain from the Nor-
man Conquest onwards.? Boards may be seen on the cloister seats of many
cathedrals of monastic origin, on the steps to the chapter house at Furness
Abbey, on tombstones at Dryburgh Abbey, Worksop, and Arbory in the
Isle of Man, at Norwich, Dover, Helmsley, and Scarborough Castles, and on
stones built into the walls of churches, as at Hargrave (Northants), Ickford
(Bucks.), Sparsholt (Berks.), and Kirkby Underdale (Yorks.).3
Combined boards, for chess, the larger merels, and tables, were often
used in Europe from the fourteenth century onwards, the board taking the
form of a shallow square box with a hinged lid. When closed, one surface
bore the chessboard and the other the merels board; when opened, the
interior showed the backgammon board. For a handsome board of this kind,
inlaid with ivory, mother of pearl, and metal, the work of Hans Sebald
Beham, 1520-40, see HC., plate facing p. 757. These boards are often men-
tioned in inventories: La Borde (op. cit.) cites:
1412. Comptes roy.: Un eschiquier de jaspre et de cristal fait aux armes de feu pape
Gregoire [1370-8 or 1406-9], et est par dehors de cippres, et y a un marellier de
marqueterie, et est garni d’eschas de mesme en un estui.
1416. Inv. Jean de Berry: Une tres belle table ployant en trois piéces, en laquelle est
le marelier, deux jeux de tables, et l’eschiquier, faiz de pourfiz de Romme. Une table
de bois marquetée du jeu des eschas, et de tables et de mareliers et sont les tresteaux
tenant a la dicte table.
Several of these box-boards may be seen in the Victoria and Albert Museum,
South Kensington, dating from the sixteenth century, e.g. nos. 4429 (1859),
7832 (1862), 220 (1870), and 154 (1900). In modern times in Italy* and Ger-
many, board F is often set out on the back of the chessboard. This is so
common that G. S. Corvinus’ defines merels thus: “Miihle ist ein Spiel auf
* R. A. S. Macalister, Archaeology ofIreland, 1928, p. 123.
? The board is drawn in MS. Trin. Coll. Cambridge, O.2.45, f. 2b, written at Cerne Abbey,
Dorset, in the second half of the thirteenth century.
= Also in the cloisters of San Paolo without the walls, Rome; and two boards were found at
Pilgrim’s Castle, Atlit, Palestine, built by Crusaders 1217-91 (C. N. Johns, Excavations at Pilgrim’s
Castle, 1932-3).
* Already mentioned in Carrera, II Gioco de gli Scacchi, 1617, 36.
5 Aramanthes, 3¢ Aufl. 1733.
 3-5-1] AND CONFIGURATION 45
dem umgekehrten Damenbrete’, and some Italian game-manuals simply
define merels as ‘the game on the back of the chessboard’.
The larger merels is frequently mentioned in the medieval literature of
France, England, and Germany. It is described fully in Alf., and the BS. and
CB. works contain collections of problems of the game. The introductory
poem in CB. says “Tunc merellos doceo quibus plebs iocatur’, showing the
popularity of the game among the commonialty. In France particularly, the
game gave rise to many metaphors of which Godefroi gives many examples;
traire la merele, traire de bonne merele, traire fausse, sauve, autre merele, &c.,
mestraire le merel, mestraire la merele, changer la merele, &c.
The game has also been played with living pieces. Such a game was
played 24 June 1897 at Saffron Walden, with boys and girls as pieces, in
connexion with Lord Winchelsea’s Order of Chivalry.
A. v. Hahn! gives the following score of a game on board F (notation as in
Fig. 18):
White Black White Black
I a7 b7 9 a2 atc (2)
2 a3 b6 10 b2-c2 a6—as
3 bs b8 Ti. aj—ad b6-a6
4 br a8 12 c2-b2: b4 b7-b6
5 c8 a6 13 b2-c2 b8-b7
6 6 C7 14 b1—b8
7 b3 b4 winning
8 bz: b4 b4
Note (1). Or
Dames c2 13 b2-a2:a5 b6-a6
10 a2-alI b4-c4 14 bs—b6 a6-a5
11 b2—a2: c2 C4-C3 I$ a3-a4 as—a6
12 a2—b2: c3 a6—a5 16 a4-a5 or a4—a3 : a6 winning

This game and its variation illustrate two positions which are often given
special names. At move 10 in the main play, White has men on a3, bi, c3, and
b3, b6 being empty, so that, if it is turn to play, he can make a row in four
different ways. This position is called krossmylna or vangjamylna in Iceland
and afarri or irman in Somaliland. At move 10 of the variation, White has
men on bi, b2, b3, a1, and a3, and a2 is empty, so that he can make and
unmake a row on every move. This position is called zwickmiihle in Germany,
rendemélle, ‘running mill’, in Denmark, svikamylna or rennihestur in Iceland,
running Jenny in Yorkshire, diporto in Macedonia, and charri or saddeh in
Somaliland.
The larger merels games are:
3-5-1. Spain: Alquerque de nueve (Alf. 92a); Real or castro in Castilian
Spanish; Marro in Catalan (Fiske, 98-104). Played on board F only. Accord-
ing to Alf. the game may be played with or without dice. The account is not
Buch der Spiele, 3e Aufl. 1900.
46 GAMES OF ALINEMENT [ois
very clear, but apparently three dice were used for the entry of men on the
board, and throws of 6, 5, 4 or 6, 3, 3 or 5, 2, 2 or 4, I, I gave the thrower
the right to enter a row of three men and to capture one, or, if another row is
produced with men already entered, two of the opponent’s men. Other
throws only enter a single man. When all the men are entered, the dice are
discarded, and the game is played in the usual way. This dice variety is not
mentioned in any later Spanish works which only describe the ordinary
method of play.
3-5-2. France: Mérelles (Fiske, 110-17); in MF. mereles or jeu de merelier;
in modern French also jeu de moulin, and dialectally, maréne. Board F only in
the Middle Ages, but Moulidars, who says that the game has undeservedly
fallen into disuse, gives board G and adds the additional rules, (1) no rows
may be made on the slant lines, (2) when a player is reduced to four men, his
men can leap to any vacant point. The Vetula (1. xxxiv) and its French transla-
tion, La Vieille, mention the dice variety given in Alf.
3-5.3- Italy: Smerelli. In ML. merelli. Also riga di noue in Carrera
(p. 36). In Hyde (p. 203) smerelli or tavola da molino. In modern manuals,
tavola, tria, filo, filetto, filo-mulino, mulinello (Fiske, 106-10). Boards
F and G. Captures can only be made from a row when no other man is
available. In the Middle Ages this was a matter for agreement before the
game began. Thus CB. 1 says: ‘Imo possunt capi item illi qui sunt in lineis
factis quam alii; et ita debes protestari quando facis partita.’
3-5-4. Britain: Nine men’s morris. Most of the perversions of merels
given above are used in dialects, and to distinguish this game the generic
term is preceded by nine men’s, nine pin, nine penny, or nine peg. Hyde (p. 204)
adds Bushels (probably from Shropshire). Usually played on board F. When
board G is used each player has eleven or twelve men. In these varieties men
may move along the slant lines but rows cannot be made on them.! Twelve
men’s morris was carried to New England by early British settlers, and is the
standard form in the United States where rows can be made on the slant
lines, and when a player is reduced to four men, his men can leap to any
empty point.
The space within the central square often contains a circle (Hyde, 204;
Yorkshire Weekly Post, 3 July 1915), and is called the bushel or pound. Captured
men are put in this circle.
3-5-5. Germany: Miihle, miihle(n)spiel. Also miilchenspiel or doppel-
miihle (Hyde, 203, has mulen and dupelmulen), and dialectally near Niirnberg,
schafzagel, ‘chess’ (Schmeller, Bairisches Worterbuch, iti. 334). Board F only
(Fiske, 127-32).
3-5-6. Holland: Molenspel (Fiske, 133); Negensticken (Hyde, 204).
3-5-7. Denmark: Mlle (Fiske, 134).
3-5-8. Sweden: Qvarn, dubbel-qvarn (Fiske, 134-8). Board F, but
™ Some modern manuals, e.g. Boy’s Own Book (1868), 630, allow rows to be made on the
slant lines.
3.5.8] AND CONFIGURATION 47
board G is sometimes used. When a player is reduced to three men on F or
four on G, his men can leap to any vacant point.
3-5-9. Iceland: Mylna (Isl. Gatur, 300-2; Fiske, 138-42). Board F only.
3.5.10. Russia: Melnitsa or Melnichny.
3-5-11. Hungary: Malomjatek, malmosdi (Fiske, 133); Czechoslovakia:
mlyn (A. G. Shirreff).
3-512. Macedonia and southern Greece: Triodi (G. F. Abbott, p. 295);
triodion (Hyde, 205). Board F and occasionally G, but with only nine men.
3.5.13. Arabs of Arabia and Palestine: Dris, riz, dris at-tan‘ashari, dris
at-tis‘a. Formerly qirq and at different periods sudder, sudra, k‘ab al-baidar,
tubn, tubna (Hyde, 205-6). Usually played on board F, but al-Firtizabadi (op.
cit.) gives board B, and K. Niebuhr (op. cit.) gives both boards. In Palestine,
captures can only be made from a row (dris) when no other man is available
(Hilmi Samara).
3.5.14. Persia: Dris, sidere, k‘ab al-baidar, si-perde (Hyde, 206).
3-5-15. Turkey: Duqurjin (Hyde, 206).
3-5-16. Armenia: Sgjoug, dugh (Hyde, 205). Board F.
3-5-17- India, Punjab: Nao-guti, ‘nine game’ (Gupta, d, 143). Board F.
Said to be similar to bara-guti pait pait, played at Vikrampur, E. Bengal.
3.5.18. India, United Provinces, Karwi Subdivision: Sujjua (Humphries,
117). Board F. The diagram of the board occurs on a slab in the wall of the
inner shrine of the ruined Chandel temple, Baldewa, close to the railway
and about two miles from Karwii station.
After each player has entered his first man, he may in following moves
either move a man already entered or enter a man.
3-5-19. Manipur, Assam; Kabui Nagas: name unrecorded (T. C. Hodson,
The Naga Tribes ofManipur, 62). Board G, but Hodson omits the lines joining
a4 and c4, a8 and c8 (see Fig. 18F), possibly because he copied his diagram
from a blurred board on the ground. Captures can only be made from a row
when no other man is available.
3.5.20. Southern India: name unrecorded. (Parker, 579 ff.)
3.5.21. Ceylon: Nerenchi keliya, niranchi (L. Ludovisi, “Sports and
Games of the Singhalese’, JRAS., Ceylon, 1873, p. 34; Parker, 507 ff.).
Board F with nine men a side, board G with twelve. During the entry of
men, a player who makes a row enters another man but apparently does not
take an opponent’s man. When all the men are entered, a player who makes
a row, takes an opponent’s man and has an extra move.
3.5.22. China: Sam k’i, ‘three game’; chuk sam, ‘taking three’ (Culin,
a, 102; Lister, Notes and Queries in China and Japan, 1870, p. 127; Himly, a,
477). Board F (siao hang) and board G (ta hang) are both used. Each player
has twelve men on each board. If a player makes a row during entry, he
places a man on top of the man he intends to take. When all the men are
entered, the ‘dead’ men are removed from the board and the game continues
in the usual way.
48 GAMES OF ALINEMENT ICH
3.5.23. Korea: Kontjil (Culin, a, 102). The Chinese game.
3-5.24. Ancient Egypt: name unrecorded. Board F at Kurna (Fig. 7).
3-5-25. Somaliland: Shah (R. G. Drake-Brockman, British Somaliland,
London, 1912, p. 129; G. Marin, a, 503). Board F and each player has twelve
men. During entry no man can be taken although rows may be made.
When entry is completed, every point is occupied. If any rows have been
made during entry, the player who was first to make a row now begins by
removing any one of his opponent’s men; if no rows have been made, the
player to enter the last man begins by taking an opponent’s man. If at any
time the player whose turn it is to play is unable to move, his opponent
makes a second move to give him a man to play, but he may not capture on
this move even if he makes a row. The winner of a game scores his win by
placing a pebble in the pound. Ifa series of games is played, the loss of a game
cancels all previous wins. Four wins in succession is called gal, ‘pool’, and
five, gabad, ‘girl’.
3.5.26. Gold Coast and Nigeria: Akidada (K. C. Murray, who saw it at
Nopa, Nigeria). Board F, drawn on the ground. Each player has nine or
ten men.
The following games of American Indians are taken from Culin(g, 793-801):
3.5.27. Keresan stock, Keres Indians, Cochiti, New Mexico: Paitariya;
Tanoan stock, Tigua Indians, Isleta, New Mexico: Picaria. Board F; the
bisectors of the sides are prolonged outside the squares.
3.5.28. Tanoan stock, Tewa Indians, Santa Clara, New Mexico: Pita-
rilla; Bidaria. The bisectors of the sides do not extend beyond the squares.
3.5.29. Zufiian stock, Zufii Indians) New Mexico: Awithlaknanai.
Board F.
3.5.30. Mariposan stock, Yokuts Indians, Tulare Cty., California: name
unrecorded. Board F and clay pawn-shaped men, black and red.
3-5.31. Shoshonian stock, Mono Indians, Madera Cty., California: Yaka-
maido. Board G, with pegged men.
3-5-32- Bogas, Amazon River: Trique (A. R. Goddard, “Nine Men’s
Morris’, in Viking Club Saga-book, Coventry 1901).
6. SHIVA. A three-in-a-row game, on the cells of a latticed board or a
rectangular or square set of holes in the ground, is played in the Sahara and
neighbouring parts of North Africa. The board is usually one of 5X6 or
6X6 cells (in one game forty cells). Two persons play, each with twelve (in
one game thirteen) men, which are entered, one by one, in alternate turns of
play. When all the men have been entered, the game proceeds by alternate
moves, a man moving one step only in one of the orthogonal directions.
When a row of three men is made the player takes one of his opponent’s
men. The player who is first reduced to two men loses. There are some varia-
tions of rule in the different games which I give below; these concern the
methods of entry and alinements of more than three men. One game is used
for divination.
3.6.1] AND CONFIGURATION 49
3-6.1. Gold Coast, north of Ashanti: Wari.! Board 6 x 6 cells. Two players,
each having twelve men which have to be entered, one by one, in alternate
turns of play. No captures can be made until all the men have been entered.
A player may enter three, four, or more men along a row or column of the
board, or move a man, after all have been entered, to produce a line of four,
but a line of four does not count as a row, and the removal of an end man
from a line of four does not make the remaining three a row nor permit a
capture.
3.6.2. Nigeria, Tiv tribe, Benue Province, Gboko: Shiva; Tiv tribe,
Bornu Province, Ngala. Kare or karnun, ‘three-in-a-row’ (K. C. Murray).
Board 5X6 or 6X6 cells. Two players, each with twelve men. Neither
player may make a row until all the men have been entered. A player when
entering his men may not aline more than three men in the orthogonal
directions. Once a row has been made, no man of that row may be moved
again. Moves can only be made in the orthogonal directions and only one
step. If a player makes three rows before his opponent makes one, he wins
the game outright.
3.6.3. Sudan, Baggara tribes: Dala or Dali (R. Davies, SNR. viii. 139
and xvi. 116). Board 6x6 cells. Two players, each with twelve ‘sticks’, one
player leaving the bark on his sticks, the other peeling his. Men are entered,
two at a time, in alternate turns of play and the four central holes (nugara,
pl. nugar) must be filled first. It is permissible to enter four men in a line in
the orthogonal directions. When all are entered the game proceeds by alter-
nate moves of one step orthogonally. When a row (ta‘na) is made, one of the
opponent’s men is taken. If a line of four men had been made during entry
the player can reduce them to three by moving an end man, and this counts
as a row and permits a capture. The position zwickmiihle (p. 45) is known
as thar.
3.6.4. Sahara, Tamacheq: Dra; Hausa: dili; Bambara, Bozo, Sourai:
wali; Peul Macina: kyoti (Th. Monod in Notes Africaines (Inst. Fr. d’Afrique
noire), Jan. 1950, no. 45, p. 12). Board of 5 x6 cells. Two players, each with
twelve men, which are entered, one by one, in alternate turns of play.
When entering, a row may not be made. When all are entered, play proceeds
by alternate moves, one step orthogonally but more than three men cannot
be alined. On making a row the player takes one of his opponent’s men. If
a move makes a row in two ways, only one capture may be made. The
position zwickmiihle is called ‘a horse’ (Tamachegq, ats; Hausa, doki). Usually
games are played for ten points, a won game scoring one and if the winner
has lost no man, two.
3-6.5. Central Sahara, Tuareg tribes: Alkarhat (F. R. Rodd, ‘People of the
Veil, London 1926). Board 6x6 cells, but Rodd quotes Jean (‘Les Touareg
du Sud-Est’ in L’ Air, Paris 1909, p. 215) as saying forty cells. Two players,
1 A.W. Cardinall, In Ashanti and Beyond, London 1927, p. 255. Cardinall says that the game
has many other names. Wari is really the name of a mancala game.
5356 B
$0 GAMES OF ALINEMENT [CH. 3
each with thirteen men. Used for divination; a holy man presides, and the
winner of three games running carries the alternative submitted for divine
decision.
FIVE-IN-A-ROW GAMES
3.7.1. Japan: Gomok-narabe (Cho-Yo, Japanese Chess, Chicago 1905,
212). Played chiefly by women and children on the cells of the i-go (Chinese
wei-k’i) board which is an unchequered lattice of 18 18 cells. I know no
Japanese description of the board, but from the next game I conclude that it
is a game, first of entry and then of movement, in which a man can only
move one step in an orthogonal direction and the row of five men must be
made orthogonally and not diagonally. The game was introduced into
Europe about 1885 and is generally known as go-bang, ‘i-go board’.
3.7.2. Belgium: Go-bang (Kraitchik, 280). Played by two persons on a
board of 16 x 16, 18 X 18, or 20 20 cells. Each player has an agreed number
of men which are entered on the cells, one at a time in alternate turns of play.
The aim is to aline five men along a row or column of the board. When all
the men are entered, men acquire the power of moving but only one step in
the orthogonal directions.
3-7-3. Britain, United States, and other English-speaking countries: Go-
bang; United States, also pegity. In England played on an unchequered
lattice of 14X14 cells. In the U.S.A. metal lattices of 16Xx16 cells, each cell
containing a hole for the insertion of pegged men (whence its registered
name). A game of entry only, the aim being to aline five men in row, column
or diagonal. Two persons play as a rule but the game can also be played by
three persons, but in this case, each player is responsible for countering the
plan of the preceding player as initiated by his last move. Each player has an
unlimited number of men, the sides being determined by colour, and in each
turn of play one man is entered on an empty cell. A man, once entered,
cannot be moved.
(Another five-in-a-row game is included in French and Italian manuals
where it is taken from French manuals. It is claimed to be the game of
petteia (2.7.1.), but this it is certainly not.)
3-7-4. France: Marelle quadruple; Italy: Mulinello quadruple (Fiske,
108, 117). Played by two persons on the alquerque board (Fig. 27). Two
persons play, each with five men, which are entered one at a time in alternate
powers of move on the points of the board, and when all are entered, the men
acquire the power of moving one step along a marked line through the point
on which they stand. The aim is to aline five men along a marked line. 1 am
doubtful whether the game has ever been played.

GAMES OF CONFIGURATION
Games of configuration are ingenious and of modern invention. They are
played by two or four persons on latticed boards. The aim for each player is
3.8.1] AND CONFIGURATION §I
to occupy the position in which his opponent’s men were originally placed.
A further complication may be introduced by differentiating the men and
requiring each man to occupy the original cell of his vis a vis. All the men
have a choice of move, (1) one step to an adjoining empty cell, in some
games in any direction, in others diagonally only, or (2) if the next cell in the

a8 Seka 2
NON WW SUS
BS
vow Sl _
SN WOW WN WW NUS
fe
Re Ne tN tf a
SSS
PW
az WW WN WOW WN
WOW WW WNW WN
WW WW WN WN WN
Nii VBA
TW WW WW WW
ree
WIS CW WW SN
BS BENGE SENES s
NNN WN NS
TW WW SN WON
CHC BNBABNE CECE FIG. 19. HALMA

permitted direction is occupied by a friend or foe, and the cell immediately

beyond is empty, a leap over the intervening man to this cell, and this leap
may be repeated indefinitely as long as farther leaps are possible. No captures
can be made.
3-8.1. Halma, ‘leap’; formerly in Britain also hoppity. This game was
invented about 1880 and has achieved popularity. Played by two persons,
each with nineteen men, or by four, each with thirteen men, on a board of
16 X 16 cells. The four-handed game requires camps of thirteen cells (four on
$2 GAMES OF ALINEMENT [CH. 3
the first and second rows, three on the third, and two on the fourth) in each
corner of the board; the two-handed game requires two camps of nineteen
cells at opposite corners (five cells on the first two rows, four on the third,
three on the fourth, and two on the fifth). The boundaries of the camp are
marked by a thicker or double line, and boards on sale usually show both
sizes of camp. Only one man may be moved in any turn of play; it may be
either a move of one step in any direction to an empty adjacent cell, or a
leap over a man on an adjoining cell in any direction to an empty cell imme-
diately beyond. The first player to occupy the opposite camp wins.
Kraitchik (pp. 310-11) gives two variants of the two-handed halma,
Grasshopper, played in the same way by two players, each with ten men
on the chessboard with camps of ten cells in opposite corners, and a similar
game on the 10 10 board, between sides of fifteen men.
3.8.2. Salta. Invented about 1901 and exhibited at the Monte Carlo
Chess Tourney the same year (Brit. Chess Mag., 1901, p. 241. It is played on
the black cells of a chequered lattice of 10x 10 cells. Played by two persons,
each with fifteen men arranged on the first three rows from each end of the
board, which is placed with ar a black cell. The men on the first row are
stars, on the second row moons, and on the third row suns, and in each row
the men are numbered 1 to 5 from left to right. These emblems confer no
difference in power of move, but serve to individualize the men. Each man
moves one step diagonally, forwards or backwards, or can leap over the
next cell, if occupied, to the cell beyond if it is empty, and can continue to
leap in the same way as long as leaps are available. The aim of the game is to
occupy all the cells originally held by the opponent, placing on each cell a
man with the same emblem and number. Thus the man on ar with one star
must be placed on j1o where the original man had a star with the same
number, and so with the other cells.
 ae
WAR-GAMES
In this chapter I deal with those games which are obviously symbolical of
different aspects of warfare, or have been so regarded by the peoples who
play them. These games are normally played by two persons who, as Groos!
has remarked, appear as leaders of opposing forces and originators of strategic
operations. The operations which are exemplified in board-games have been
listed in Chapter 1? and I begin with the battle.

BATTLE-GAMES
The typical battle-game is one in which two players direct a conflict
between two armies of equal strength upon a field of battle, circumscribed in
extent and offering no advantage of ground to either army. In the older
games the board is usually an unchequered lattice of considerable size, and,
at first, players were not particular as to its dimensions. It took time and
experience to realize that the 8 x 8 board was the optimum, large enough to
give ample scope for manceuvring and not too large to maintain its value as
a means of recreation.
In the oldest, and also in the majority of existing battle-games, there is no
differentiation of man beyond that necessary to distinguish the two sides, and
there is no necessity for promotion. When later the men were restricted to
moves in the forward direction, and so lost all power of movement on reach-
ing the opposite edge of the board, this drawback was remedied by giving
promotion to any man that reached the opposite edge, with a different power
of move in order to restore its mobility.
I classify the battle-games by, first, the method of capture, beginning with
those that employ the interception capture—the oldest form of capture in
war-games—and, second, by the type of move employed.

I. GAMES WITH THEINTERCEPTION CAPTURE AND ORTHOGONAL

MOVES

4.1.1. Ancient Greece: Petteia; Ancient Rome: Latrunculi. These,

together with their apparented games in Ancient Egypt and Palestine, have
been dealt with in Chapter 2.
4.1.2. Wales: Gwyddbwyll; Ireland: Fidchell. These have also been
treated in Chapter 2.
' Play of Man, London 1901, p. 190.
2 There is a fifth operation which is exemplified in games, the siege, but siege-games are a
later development of the hunt-games to be described in Chapter 5, and are played on the same
boards. I shall accordingly deal with them in the next chapter.
54 WAR-GAMES [CH. 4
4.1.3. Persia. Nard. The only mention of this game occurs in Firdawsi’s
Shahnama (vv. 2697-898) in his version of the story of the introduction of
chess into Persia. In the older version of this story, Nishirw4n, the King of
Persia, sent the game of nard (a derivative of the Roman race-game, alea;
see 2.8.3) to the Indian king with a challenge to discover how it was played.
Firdawsi, however, without altering the name of the game, substitutes
another game for nard, and gives the following description of it:
He made two dice of ivory with figures the colour of ebony. He then arranged an
army similar to that of chess; he placed the two sides in order of battle and distributed
the troops, ready for battle and for the assault of the town, among eight houses. The
field was black, the battle-field square, and there were two powerful kings of good
disposition who should both move without ever receiving injury. Each had at his side
an army in its arrangements, collected at the head of the field, and ready for the fray.
The two kings advanced upon the field of battle, their troops moved on all sides
around them, each trying to outgo the other; now they fought on the heights, now
on the plains; when two on one side had surprised a man by himself, he was lost to
his side, and the two armies remained face to face until it was seen who was beaten.

The game was played by two persons on a board of 8 x8 cells, and each
player had a king and eight men drawn up on the outer rows. The kings were
invulnerable; the men were captured by interception. It certainly looks like
a modification of latrunculi which may, as in Britain, have been introduced
by Roman legionaries who were guarding the frontier of the Empire.
4.1.4. Siam: Mak-yek (Capt. Low, 382). Played on the 8x8 board by
two players, each having sixteen men arranged on the cells of his first and
third rows. The man has the same move as the rook in chess, and captures
are made both by interception and intervention.
4.1.5. Malaya: Apit-sodok (R. J. Wilkinson, 57, who calls it apit. I owe
the complete name to W. W. Skeat). The same game as 4.1.4.
4.1.6. Japan: Hasami-shogi, ‘intercepting chess’ (Prof. Tsuboi). Two
persons play, each with nine men arranged on his first row of the Japanese
chessboard of 9x9 cells. The chess pawns (fu) are used as men; they have
the move of the rook in chess, and capture by the interception method.
Professor Tsuboi thought that it was a modern game, a simplification of
the Japanese chess.
4.1.7. Egypt: Siga, seega (Lane, 320-1; Falkener, 63-67, quoting an
article by Dr. H. C. Bolton in the Field, 1 June 1889; Parker, 603-4). Played
by two persons on boards of 5 x 5, 7X7, or 9X9 holes made on the ground
or on the cells of latticed boards. Several such boards have been scratched by
Arab guides on the summit of the Great Pyramid and on fallen roofing-
slabs at Kurna, where these boards are markedly inferior to the carefully cut
boards made by the masons who originally shaped the slabs. They are not
very old, and Hyde did not know the game.
Both players have the same number of men (kelb, ‘dog’, pl. kilab), the total
number being sufficient to fill all the cells except the central one. The men
4.1.7] WAR-GAMES 55

are entered, two at a time, in alternate turns of play, but the central cell can
only be used after all the men are entered. When all the men are entered,
play continues by alternate moves of one man only, a move being one step
in an orthogonal direction. If a player, when it is his turn to move, has no
move that he can make, his opponent must provide him with a move by
removing one of his own men from the board. Captures are made by inter-
ception, and if further captures are now possible, the player continues to
make them.
4.1.8. Somaliland, Issaq tribe: Shantarad; Darod tribe: Bub (G. Marin,
a, 05). Two players. Board of 5 x 5 cells. Each player has twelve men which
are entered, two at a time, in alternate turns of play, the central cell (deh)
being left empty. When all the men are entered, the player who entered the
last two men begins. Men move one step in an orthogonal direction, and a
move continues as long as captures can be made each time. If a player can
take men in two or three directions by a move of the same man, he takes all
of these. Captures are made by interceptions, but a man on the central cell
cannot be taken. If a player is unable to move when it is his turn to do so,
his opponent makes another move to allow him to play.
4.1.9. Sudan: Sija (R. Davies, SNR. viii. 138). Board of 5x5 cells. The
same game as 4.1.8, as are probably the next two games.
4.1.10. Hanzoan: Ufuba wa hulana (Hyde, 233). Board of 5x5 cells;
each player has twelve men.
4.1.11. Nubia: Syredgé (R. and S. Percy, 837). Board of 7x7 cells.
Small pellets of camel and goat dung are used as men.
4.1.12. Celebes: Gala (Kaudern, 310, quoting B. F. Matthes, Makassarsch-
Hollandsch Woordenboek, 1859 and Ethnographische Atlas, 1859). Kaudern
describes this game as a kind of backgammon, and the board is used else-
where for a race-game, but there is no mention of dice in Matthes’s account.
Two players. Board of 7x7 cells on which the central cell (soelisavigke)
and the middle cell of each edge-row are crosscut. Played with ten black
and thirteen white men which move in the orthogonal directions only, and
capture by interception. Black plays first and enters a man on the central
cell. The remaining men are entered, one at a time, in alternate turns of play,
each player entering his men on his own half of the board. The situation
when one player is hemmed in and unable to move, while his opponent can
still move without releasing any of the other player’s men, is called péle by
the Boegis and bdttoe-mi nai by the Macassars.
4.1.13. Scandinavia and Iceland: Tafl, hnefatafl (HC. 445); Pre-conquest
England: Tafl; Wales: Tawlbwrdd (F. R. Lewis); Ireland: name un-
recorded; Lapland: Tablut (J. E. Smith, ii. 55-58). Played on the points of
a lattice of 18 x 18 cells, or on the cells of boards of 13 X13, IIX II, 9X9, or
7x7 cells. Two persons play, one having a king, placed on the central point or
cell, and a number of men who are arranged symmetrically around the king;
the other has double the number of men who are arranged symmetrically
56 WAR-GAMES [CH. 4
round the edge of the board. Both king and men possess the rook’s move
in chess. Men are captured by interception, the central cell counting for
this purpose as occupied by the side making the capture; the king is only
captured if the four adjacent cells in row and column are all occupied by
enemy men. A man can move to a cell between two enemy men without
capture. The player with the king wins if in his turn of play the king has
an open row or column to the edge of the board; his opponent wins if he
captures the king."
Tafl (pronounced tabl) was the older, and hnefatafl the later name of a
board-game which was already played by the Scandinavian peoples before
A.D. 400. It was carried by the Norsemen to Iceland, Britain, and Ireland,
and spread to Wales. It was the only board-game played by the Saxons.
After the introduction of chess into England in the eleventh, and Scandinavia
in the twelfth century, hnefatafl fell out of use except in remote and isolated
districts: the last mentions of the game as still played are from Wales, 1587,
and Lapland, 1732.
Hnefatafl had long been forgotten when scholars began to study the older
Icelandic literature, and many guesses were made as to its identity with other
games. Fiske (pp. 69, 143, 156, 356) had no difficulty in disproving all these
guesses, though he failed to recover the game. It was my chance discovery of
an account of the Lapp game by Linnaeus that gave the clue, and the descrip-
tion of hnefatafl which I gave in my History of Chess has been confirmed by
later discoveries.
The Germanic peoples adopted the Latin word tabula in the first or second
centuries A.D. It might have been adopted in one of two senses, either as
meaning a game-board or as the name of a particular game, the Roman alea
or tabula (2.8.3) which had been seen in intercourse with Latin-speaking
peoples. That it was adopted in the first of these senses seems certain, for there
is no evidence that any Germanic people played any game on the alea or
backgammon board before the eleventh and twelfth centuries, and by then
tafl, tefl, and zabel were in common use as meaning a board-game in Scandi-
navian, Old English, and Middle High German respectively. That the Germa-
nic peoples found it necessary to adopt a foreign term for the game-board
seems to indicate that they had no board-games of their own.
It was the introduction of Christianity which brought a knowledge of the
Roman game of alea or tabula to the Germanic peoples through the eccle-
siastical canons prohibiting the play of this game by both clergy and laity.
When glosses were added to texts to explain unfamiliar terms for the un-
learned, alea in the canons was often glossed as meaning a kind of board-game,
using the various forms that the word tabula had taken, and when glosses
were collected in glossaries, alea was naturally included. Thus we find in

? This game has more affinity with a siege than with a battle. It is included here because those
who played it clearly regarded it as a war-game. The actual arrangement of the men on two ofthe
boards is shown in Figs. 25 and 26.
| WAR-GAMES $7

OHG. glossaries of the tenth century (A. v.d. Linde, pp. 55-56), ludere tabulis
= spillone zaplis and alea= zabel, and in OE. glossaries, in the Epinal
Glossary (ante 700) alea = teblas, in the Erfurt Glossary (ante 800), alea =
tefil, and in Aelfric’s Vocabulary (c. 1000), alea = tzfel; aleae = tzfelstanas;
aleator = teflere; pirgus = cyningstan on tefle; tessere vel lepuscula = federscite
teefel.
These entries in glossaries and vocabularies have been the occasion of much
misunderstanding. Thus, OE. tzfl has been defined as ‘a die for playing with’
and tzfelstan as ‘a die’, although the termination stan ‘stone’, can only mean
a game-piece. These definitions are assuredly wrong: they are due, first, to
an assumption that the glossists were defining tzfl, whereas they were
attempting to explain alea as briefly as possible for the benefit of English
readers, and, second, to a misapprehension as to the real meaning of alea. As
we have seen above (p. 29), alea was the name of a board-game: it was never
used in the classical period as an equivalent to tessera, ‘the die’, and it is very
doubtful if it was ever so used in the pre-classical period. The Epinal and
Erfurt Glossaries simply tell the reader that alea was the name of a board-
game; Aelfric, however, is clearly thinking of a particular board-game
played with a king-piece (cyningstan) and tablemen (tefelstanas).!
So long as the Scandinavian, English, and German people only played one
board-game, there was no reason why it should need a special name; tafl,
tefl, and zabel could only mean that game. But when other board-games
became known, this was no longer the case, and it was necessary to differen-
tiate between the board-games. In Iceland chess and tables became respec-
tively skaktafl and kvatrutafl; in Germany chess was schachzabel and tables
wurfzabel.
The earliest evidence that a board-game was actually played by any Ger-
manic people comes from Scandinavia. Montelius divides the first millennium
A.D. in Scandinavia into three periods: I. The Roman Iron Age, ending
c. 400; II. the Vilkerwanderungen, c. 400-800; and III. the Viking Age, c. 850-
1050. The excavation of grave-mounds of all three periods has produced
many game-materials. These include dice, both long and cubic, made of
bone, and sometimes a burial includes a set of three dice, as were customarily
used by the Romans. They also include game-pieces; in the first two periods
they are usually button-shaped with smooth base and slightly curved upper
surface, but in the second period they tend to be taller: an Ultuna mound
produced thirty-six pieces with bases, 1} in. in diameter and 1 in. in height.
In the Viking Age these sometimes bear a pattern of lines and occasionally
a man’s head. Other pieces of this Age are pawn-shaped with a knob, and
sometimes one of the pieces is differentiated, having the form of a crowned

1 Tt is immaterial that the Latin equivalents are wrong: dictionaries are often wrong about the
terms used in games. The Promptorium Parvulorum (EETS., 472) has “Tablere or a table of play or
game: pirgus’, when pyrgus means the dice-box for shaking the dice (which was never used in
Europe during the Middle Ages).
58 WAR-GAMES (CH. 4
king or an animal. These pieces are made of bone, glass, amber, and occa-
sionally, clay.
There are also several game-pieces
from Saxon and Scandinavian barrows
and cemeteries among the Anglo-Saxon
antiquities in the British Museum (H. A.
Smith, 48, 65, 79, 82, 167). Some from
Pensthorpe, Norfolk, are button-shaped
like the counters used by Roman players
for latrunculi, others from Taplow, Faver-
FIG. 20. GAME- PIECES sham, and Basingstoke are short, hollow
(after Montelius) cylinders, made from horses’ teeth with
the opposite ends closed by disks united
by a silver pin. All these may have been used for hnefatafl.
More interesting are fragments of two boards, one found at Wimose,
Fiinen, in a grave of the Roman Iron Age, the other in the Gokstad ship of
the Viking Age. Both fragments are of sufficient size to show that the com-
plete boards were square, and how they were divided. The Wimose board

(Ge FIG. 21. WIMOSE BOARD

22 eee
a cs a
eTial 2 oon oe 8
ee

FIG. 22. BOARD FROM GOKSTAD SHIP

was about 18 in. square, and each row and column contained eighteen cells,
all unchequered. The reverse of this board shows a pattern of larger and
smaller circles round the edges. The Gokstad board has on one side a square
board of 13 X13 cells, and the odd-numbered cells on the second and fifth
columns from either side of the board are chequered. On the reverse of the
board, the board F (Fig. 18) for the larger merels is set out.t
? Similar combinations of boards are mentioned in Icelandic sagas; thus, in Kréka-Refs Saga
(14th c., but based on older works), K6benhavn 1883, p. 23, Gunnar sent from Greenland to
King Harald Fairhair a tanntafl (a board of walrus bone), to which the later writer adds ‘it was
both a hnefatafl and a skaktafl’.
Astaro WAR-GAMES 59

In 1932 a wooden board, 9} in. square with 7 x7 holes for the insertion of
pegged men, was found when excavating a crannog or lake-dwelling at

i ¢,
ee
fs i j 22D 22°

prem

<

|mee
os

Se
4
ay)
i

Ng
EEO
SU
3

Se

@ Soran
—

=A
See
ty
7

a \ ip

12 vi) e oesig, (2
\\r
“YA yy

yeeMAI? | peace en
Maaintfinlg ar
: iat
Sol ae, BS"
& VRE RPRE-

Yh
Aireea i
mi)
RAY
z 2X SR

3 Bowes SASS
PRN DS ? —
. ios
= cl

FIG, 23. THE BALLINDERRY GAME-BOARD. (National Museum ofIreland)

Ballinderry near Moate, co. West Meath, Ireland. The central hole is
surrounded by a circle and the four corner holes by quadrants, and the
whole board has a frame with various tenth-century patterns which point to
60 WAR-GAMES [CH. 4
manufacture in the Isle of Man (The Times, 7 Oct. 1932, and Journal Manx
Museum, Mar. 1934, pp. 164-5). The board has two handles, carved as heads,
by which the board was held by the players.
A similar board was drawn on one of the rings of a gold horn found in
1639 at Gallehuus, South Jutland, which was stolen in 1802 from the Royal
Cabinet, Copenhagen and melted down. Fortunately a drawing of the horn
had previously been made. The third ring contained a scene in which two
men hold a board, around the edge of which sixteen small disks were regu-
larly spaced (Fiske, 309).
All these boards belong to the game of hnefatafl, and I think that the so-
called chessmen of jet found during the excavation of the Mote Hill at

FIG. 24. HNEFATAFL PIECES

A, B.Jet pieces found at Warrington

C. Bone piece from Woodperry, Oxfordshire

Warrington (Nat. Hist. Soc. Lancs. and Cheshire, v. (1852) 59) and the bone
piece found at Woodperry, Oxfordshire (Arch. Journal, iti. (1846) 121)
also belong to hnefatafl.
I now turn to literary sources.
Iceland. In the older sources the game is always tafl. Two kinds of piece are
mentioned, the hnefi (meaning doubtful, but used of the king-piece) and
hunn, ‘knob’ (used for the man).
Tafl is mentioned in the oldest poems. In Volo-spa (CPB., i. 194, written
ante 1064), the Anses are described as playing in their court and finding
golden tzflor, ‘table-men’, in the grass, which had belonged to them in days
of yore. Earl Rognwald of Orkney, the crusader (c. 1125) begins the list of
his accomplishments with ‘I am strong at tafl-play’ (ibid. ii. 276). The Rigs-
pula (ante 1200; ibid. i. 166) speaks of children learning to swim and play
tafl. Hornklofi’s Raven Song (ibid. i. 257) has ‘they are well cared for, the
warriors who move the hunns in Harald’s court’. The Greenland Lay of Atli
(ibid. i. 342) says that the hnefi is often beaten when the hunns are taken.
More illuminating are two riddles which were proposed in a contest of wits
between King Heidrek and Odin in disguise (Herverar Saga; ibid. i. 87-92):
4.1.13] WAR-GAMES 61
“Who are the maids that fight weaponless around their lord, the brown
ever sheltering and the fair ever attacking him?’ (Answer: the pieces in
hnefatafl), and “What is that beast all girdled with iron which kills the
flocks? It has eight horns but no head’ (Answer: the hnefi or head-piece in
hnefatafl).*
The only one of the many references to tafl in the sagas, to give any infor-
mation about the game, occurs in Frid pjof’s Saga (Englished in E. Magnusson
and W. Morris, Three Northern Love Stories, 1875, p. 73). Hilding is sent to
obtain advice on procedure in a war and finds Frithiof playing hnefatafl with
Bjérn. Frithiof makes no direct reply, but his answer is concealed in his
comments on the game: ‘a bare place in your board which you cannot cover,
and I shall beset your red pieces there’, and when Bjérn exclaims ‘A double
game, and two ways of meeting your play’, Frithiof replies: “Your game is
first to attack the hnefi and the double game is sure to be.” Many modern
recensions of the saga exist, but nearly all transfer the game to chess (Fiske,
25-32).
Britain. The Icelandic saga, Fornaldar Ségur (c. 1256, but incorporating
earlier material), says that Hvitserkr and Sugurd were playing hnefatafl during
a raid on Northumbria when a messenger from King Alla (died 856) arrived.
An English or Irish manuscript (CCC., Oxford, 122, see J. Armitage
Robinson, 69 ff.) contains a curious attempt to give a scriptural meaning to
hnefatafl which is here named alea. The text begins: ‘Incipit alea euangelii
quam Dubinsi, episcopus bennchorensis, detulit a rege anglorum, id est a
domu Adalstani, regis anglorum, depicta a quodam francone et a romano
sapienti, id est Isrl.’ (Akthelstan reigned 925-40, and according to the Four
Masters, Duibhinnsi, sage and bishop of the family of Beinchair, died in 953.)
The manuscript uses the board and the arrangement of the pieces as a
scheme for setting out a harmony of the Gospels, and this board is identical
in its dimensions with the Wimose board. The game is played between two
sides, one of forty-eight men, and the other of twenty-four men and a hneft.
The men are placed on the points of the board, so that there is a central point
which is occupied by the hnefi (in the allegory, primarius vir, to show the
unity of the Trinity). His twenty-four defenders are scattered symmetrically
in the middle of the board, and the forty-eight attackers are scattered sym-
metrically on the points nearer the edges. The positions of the men are used
to set out the details of the harmony. There are some errors in the diagram
of the manuscript, but these can easily be rectified, and give the arrangement
as shown in Fig. 25. The allegory is very artificial—the board contains
18X18 cells because the total of the four Evangelists, four Gospels, and ten
Canons is eighteen; there are seventy-two men because the number of items
in the harmony is seventy-two—but this need not concern us here. The
important thing is that we have a picture of hnefatafl as played in the tenth
century, which is consistent with both older and later descriptions.
* At a much later date, rhymed answers were added which substitute chess for fafl.
62 WAR-GAMES [CH. 4
I think that the passage in Gaimar’s Lestorie des Engles (Rolls Series, lines
3655-8): ‘Orgar iuout a vn esches, Vn giu kil aprist des Daneis; Od lui iuout
Elstruet la bele, Sur ciel nout donc tele damesele’, really refers to hnefatafl,

FIG, 25. ALEA EVANGELII

and is one of the many examples of the substitution of chess for an older
game.
Wales, We meet in Welsh translations of English romances, e.g. in Bown
o Hamtwn and Y Seint Greal with a board-game called tawlbwrdd (tawlbwrd,
dalwbwrd), and references to this game occur frequently in the Ancient Laws
of Wales (ed. 1841, pp. 12, 16, 143, 149, 436, 542), ascribed to Howell Dda,
but not older than 1250. An elaborate calculation of the worth of the king’s
tawlbort (p. 436) shows that the game was played between two sides, one of
sixteen men, the other of a king (brenhin) and eight men (werin), which
establishes the identity of the game with hnefatafl.
4.1.13] WAR-GAMES 63
A Peniarth manuscript (Welsh Nat. Library, 158, p. 4)! contains a descrip-
tion of tawlbwrdd which was written by Robert ap Ifan in August 1587, with
a drawing of the board. It contains 11 x 11 cells and the second, fourth, sixth,
and eighth columns are shaded (cf. the Gokstad board), and continues:
The above board must be played with a king (brenin) in the centre and twelve men
in the places next to him, and twenty-four lie in wait to capture him. These are placed,
six in the centre of every end of the board and in the six central places. Two players
move the pieces, and if one belonging to the king comes between the attackers, he
is dead and is thrown out of the play; and if one of the attackers comes between two
of the king’s men, the same. If the king himself comes between two of the attackers
and if you say ‘watch your king’ before he moves into that place, and he is unable to
escape, you catch him. If the other says gwrheill (2) and goes between two, there is no
harm. If the king can go along the line (lacuna here) that side wins the game.
Lapland. Linnaeus’s account of tablut is given by Sir J. E. Smith (ii. 55)
under date of 21 July 1732. The board (Fig. 26) has 9x9 cells, and
certain cells are chequered in different ways, (1) the central cell (konakis);

a Rc ae
eteLae
LL BEES Wf

Sopp.
VESVehe ze

FIG. 26. TABLUT, LAPLAND

(2) the orthogonally adjacent cells and the four cells beyond these; (3) the
three middle cells of each edge and the four cells completing the serified
cross on the board (cf. the rune in Finn Magnusson’s manuscript, Bodleian
93, to be held in the hand as a charm to secure a win at chess, originally, I
suppose, at hnefatafl). The pieces are a king, shaped like a chess king, eight
Swedes, pawn-shaped with a pyramidal head, and sixteen Muscovites, pawn-
shaped but with two heads. All the pieces have the move of the chess rook.
The king alone can enter the konakis. The player with the king wins, if the
1 Tam indebted to Dr. F. R. Lewis for this reference.
64 WAR-GAMES (CH. 4
king reaches an edge of the board. When he sees that his king has a clear
road to an edge, he must say raichi, and when he has two clear roads, tuichu
(cf. Frithiof’s Saga). A man is taken when the opponent occupies both adjacent
cells in row or column, but a man can move to a cell between two cells in
row or column which are already occupied by opponent men without being
taken. The king is captured if all four adjacent cells in row and column are
occupied by opponent men, and also when one adjacent cell is the konakis
and the other three on row and column are held by the enemy. These rules
are again consistent with all that we know from other sources.

I. GAMES WITH THE INTERCEPTION CAPTURES, AND DIAGONAL

MOVES
4.1.14. American Indians, Zun stock, New Mexico: Awithlaknakwe,
‘stone warriors’ (Culin, d, 877; g, 799). Played by two or four persons on a
board, a-te-a-lan-e, ‘stone plain’, of 12 x 12 cells with six additional cells along
and without the middle of each side. The diagonals of all cells are drawn; the
orthogonal lines are called a-kwi-we, ‘canyons’, and the slant lines, o-na-we,
‘trails’. When two play, North opposes South; when four play, North and
West oppose South and East. Each player has six men on the six cells outside
the square and nearest to him. The men can only move one step forwards
along the trails, i.e. diagonally forwards. Captures are made by interception,
the three men engaged all standing on the same diagonal. The first man on
each side to be taken is then replaced by a seventh man, pi-thlan-shi-wani,
‘priest of the bow’, which, in addition to the ordinary move, can cross the
canyons, i.e. move one step orthogonally. No piece may move backwards.
F. H. Cushing (who collected this game) says that the aim of the players is
to cross over and take the opponent’s place, capturing as many of his men as
is possible on the way.
The men used in this game are disks, about one inch in diameter, one
player having plain disks and the other disks with a small hole bored through
their centre. They are made from broken fragments of earthenware vessels.
The game is commonly played upon housetops which are often marked
with the diagram of the board. Pottery disks are frequently found among the
ruins of houses, and specimens from the cliff dwellings in the Mancos Canyon,
Colorado, are in the Museum of Archaeology, University of Pennsylvania.
Similar disks have been found in many parts of North and South America,
in mounds on the Georgia coast, on the sites of ancient villages in Ontario,
Canada, east and south-east of Lake Simene, in Peru, Bolivia, and Argentina.
These suggest that this or a similar game has been widely played and is of
considerable antiquity.
4.1.15. American Indians, Moki stock, New Mexico: Totoldspi (Culin,
d, 879; g, 795). Played on the points of a latticed board of 10 10 cells on
which the dexter diagonal (known as tu-ki-o-ta) is drawn. Two or more
persons play, and sit at either end of the diagonal, moving their men along
4.1.15] WAR-GAMES 65

the diagonal in opposite directions. Each player has fifty-five men which are
placed on the points of the board, one player above, and the other below the
diagonal. Culin’s description is obscure and incomplete, but it appears that
all moves take place along the diagonal only, so that the moves are neces-
sarily diagonal, that certain positions entitle the player to move more than
one man in the same turn of play, that captures are made by interception,
and a move may include several captures consecutively. He describes the
play as ‘intricate’ and adds that the game was only rarely played on the East
Mesa, but was still played at Oraibi in 1898. It is said to have been played by
the sun and moon and other mythical personages. Boards, carved on the
rocks near the village of Walpi, have been noted.

2. GAMES WITH THE LEAP CAPTURE, AND MOVES ALONG THE

MARKED LINES OF THEIR BOARDS

Alquerque is the older name of a battle-board which was introduced into

Spain by the Moors, and I propose to use this name for a group of battle-
games which are played on lined boards that have been developed from the
board on which the smaller merels is played (see p. 6). The primitive board
of the alquerque games is shown in Fig. 27; it contains twenty-five points on
which the men are arranged, and the lines of
the board show the directions in which moves
can be made from each point. The game, from
its diffusion, appears to be an old one, and I
have suggested that an unfinished board at
Kurna (Fig. 7D) may have been intended for
this game. Two persons play, each with the
same number of men, and play starts from a
given arrangement. A man, stationed on any
point, can move along any marked line through
that point to the next point on that line, which FIG. 27. ALQUERQUE
must be empty. If the adjacent point is occupied
by an opponent’s man and the point next beyond it on the line is empty,
then the man can leap over the enemy man to the vacant point beyond and
take the enemy man by a short leap. If now another enemy man is en prise,
it can be taken by a second short leap, and further captures may be made in
the same turn of play, so long as an enemy man is en prise after each capture.
In some games, it 1s compulsory to take.
4.2.1. Spain: Alquerque de doze (Alf. 91a); Alquerque (Covarrubias;
Brunet y Bellet, 204-6). Played with twelve men on each side, arranged as in
Fig. 27. According to Covarrubias, capture is compulsory, and any failure to
take is penalized by the huff.
4.2.2. Catalonia: Marro (Brunet y Bellet).
4.2.3. Italy: Marelle (Brunet y Bellet); Sicily: Marella; Riga (Carrera,
5356 F
66 W AR-GAMES [CH. 4
33-36). The board is incised in stone in the cloisters of San Paolo without
the walls, Rome (Miss K. M. E. Murray).
4.2.4. France: ‘jeu de mereles qui se fait par douze mereles’ (La Vieille,
lines 1742-6).
4.2.5. Switzerland: name unrecorded. The board is set out on one of the
outer surfaces of a game-box, late sixteenth century, which formerly belonged
to the Wappenwyll family, and is now in the Victoria and Albert Museum,
South Kensington, no. 96 (1908).
4.2.6. England: name unrecorded. The board (arrangement of Fig. 27,
reflected) in MS. Trin. Coll., Cambridge, 0.2.45, £. 2b. A board has been
drawn on the chancel tomb of Sir George Colt (died 1570) in St. Mary’s,
Cavendish, Suffolk, and another, much worn, is in the cloisters of Norwich
Cathedral.
4.2.7. India, Punjab: Bara-guti (H. C. D. Gupta, a, 570). The arrange-
ment of the men in Fig. 27 is reflected.
4.2.8. India, United Provinces, Moradabad, and Naini Tal: name un-
recorded (seen by A. G. Sherriff).
4.2.9. Assam, Naga tribes: name unrecorded (T. C. Hodson, 62). The
board was drawn on the ground and Hodson, who gives two diagrams of the
board, omits the slant lines in one and the second and fourth rows in the other.
4.2.10. Assam, Sema Naga tribes: name unrecorded (J. H. Hutton, b, 111).
Eleven men on each side, the two inner men on the middle row (Fig. 27) are
omitted.
4.2.11. Assam, Angami Naga tribes: Terhiichii (J. H. Hutton, a, 101-2).
Ten men on each side, all the men on the middle row being omitted. Some-
times also played with eight men on each side, omitting in addition the men
on the outside points of the second and fourth rows.
4.2.12. Palestine: Natt klab ash-shawk (Hilmi Samara). Played on the
cells of a 5x5 board with no slant lines, the central cell being cross cut.
Twelve men arranged as in Fig. 27. The men can only move one step ortho-
gonally, and captures are made by the short leap.
4.2.13. Abyssinia: name unrecorded (Griaule, a, 176). Seen at Choa near
Gondar and at Ankobar (where it is played by Catholic students). Board and
arrangement as in Fig. 27. A player loses when he is reduced to five men.
Also played at Gondar on a board which omits all slant lines, presumably
with orthogonal moves of one step only. It is said to have been introduced
by the soldiers of K. Walda Giorgis.
4.2.14. Sahara: Damma, zamma (Monod, 12). Board and arrange-
ment as in Fig. 27. The board is called kharberg. The standard game with
compulsory captures.
4.2.15. Madagascar: name unrecorded (Montgomery, 148-56; Dubois,
844). No longer played, having been replaced by Fanorona about 1680.
The American Indian tribes have learnt the game from Spanish settlers in
Mexico.
4.2.16] WAR-GAMES 67
4.2.16. Keresean stock, Keres Indians, Acoma, New Mexico: Aiyawat-
stani, ‘chuck away grains’ (Culin, g, 792). The two outside men on the
middle row of Fig. 27 are interchanged: this is the case in the other American
Indian games with sides of twelve men.
4.2.17. Zufian stock, Zufi Indians, New Mexico: Awithlaknanai
(Culin, g, 800). The outer edge of the board is.a circle. The game is said to
be a Mexican game.
4.2.18. Piman stock, Papago Indians, Pima County, Arizona: Pon
chochotl, ‘coyote and chickens’ (Culin, g, 794).
4.2.19. Shoshonean stock, Hopi Indians, Oraibi, Arizona: Tuknana-
vuhpi (Culin, g, 794). A stone board is in the Free Museum of Science and
Art, University of Pennsylvania (no. 38613). The central point is marked
with a star and all the diagonal lines are drawn. Played
intwo ways: (1) Twelve
men on. each side, arranged as in 4.2.14. (2) Ten men on each side arranged
on the two outer rows at each end of the board. In both varieties, whenever
a row or column becomes empty, it is no longer used.
In parts of Asia two triangles are added to the board of Fig. 27 and each
player has sixteen men arranged as in Fig. 28, or eighteen, adding the extra
men on the middle row as in Fig. 27. The
rules of move and capture are unchanged. This
enlarged board (sometimes called “Double

NAINA
Alquerque’) is used in the following games:
4.2.20. Turkestan: Shanzdahu-kattaru

ZAIN ZEISS
(Persian), Uno-alti-tashu (Oboikh), ‘sixteen
men’ (P. Komarov, 250). The triangles are

SDS
called astar in Persian, krepost in Oboikh.
4.2.21. India, Bengal: Solah guttiya, ‘six-

YINZN
teen balls’ (Parker, 583); Lower Bengal:
Mughal-Pathan (J. M. Datta, 167).
4.2.22. India, United Provinces, Karwi
Subdivision: Athara guti, also baz, mar,
ticcha or bangale (E. de M. Humphries, 117).
Far more played than pachisi (6.4.6). Athara ric, 28. MOGOL PUTT’HAN
gutiala teoria (H. C. D. Gupta, c, 165) in the
Central Provinces. Mangal-pata (H. C. D. Gupta, a, 570) in East Bengal
and Behar. Each player has sixteen men.
4.2.23. Assam and Sikkim: Dam pusri or Sipchi kat (S. L. Hora, 2).
Each player has eighteen men.
4.2.24. Manipur, Assam: name unrecorded (T. C. Hodson, 62). Sixteen
men on each side.
4.2.25. Deccan: Mogol Putt’han (G. A. Herklots, App. vii). Each player
has sixteen men.
4.2.26. Ceylon: Hewakam keliya, ‘war-game’ (Parker, 583). Each
player has sixteen men.
68 WAR-GAMES [CH. 4
4.2.27. Malaya: Dam (K. Plitschke, 191). Each player has sixteen men.
The triangles are called gunung, ‘mountains’ and the central point pusat,
‘navel’.
4.2.28. Indonesia, Saleijer: name unrecorded (H. E. D. Engelhards, 315).
Each player has sixteen men.
Two games with the same rules of move and capture are played in Ceylon
with an increased number of men on the board of Fig. 29.

FIG. 29. PERALI KOTUMA FIG. 30. TERHUCHU

4.2.29. Ceylon: Perali kotuma, ‘war enclosure’ (Parker, 583, who

says that this game is also played in India). Each player has twenty-
three men.
4.2.30. Ceylon: Kotu ellima (L. Ludovisi, i. 34; Parker, 583). Each
player has twenty-four men, the extra man being placed on the middle row
between the player’s triangle (that to his left hand) and the central point.
A more complicated board is used by the Angami Naga tribes of Assam
on the board of Fig. 30.
4.2.31. Assam, Angami Nagas: Terhiichii (J. H. Hutton, a, 102). Played
between sides of nine men arranged as in Fig. 30. The eight triangles are
refuges to which men can retire. A man within a triangle can skip two points
along any line, e.g. from corner to corner.
The quadruple alquerque board which occurs on the roofing slabs at
Kurna (Fig. 7G) is used for an alquerque game in India, Indonesia, and the
Sahara with sides of forty men arranged as in Fig. 31. In Asia the rules of
move and capture are unchanged and captures are not compulsory.
4.2.32. India, Punjab: Ratti-chitti-bakri (H. C. D. Gupta, d, 143).
4.2.33. Achehn, Sumatra: Meurimiieng-rimiieng peiiet ploh, ‘tiger
game played with forty’ (Snouck-Hurgronje, I. iii. 204).
4.2.34. Java: Dandaman, dam-daman (Sir T. S. Raffles, i. 350; D. M.
Campbell, ii. 1017).
4.2.35] WAR-GAMES 69
4.2.35. Sahara: Damma, zamma (Monod, 11). The board differs from
that of Fig. 31 in that the second, fourth, sixth, and eighth lines in row and
column are omitted, thus reducing the number of lines through points on
these rows and columns, and the powers of move from these points. The

WANA BS
oeANid
SUAS
LASVA
ESLAN
ANNA
easIN4)Bd
[2BS
4 NAN
VASE
LIN
NZWINNZ NIVAN
a - JINY
a, \ th NZIS
NIINZINVIN Wa FIG. 31. QUADRUPLE ALQUERQUE

sides are arranged as in Fig. 31. The black men are ‘males’ and usually small
bits of stalk; the white are ‘females’ and usually pellets of camel-dung. The
player with the ‘males’ always plays first. The man moves one step at a time
and only in the forward directions along marked lines; but takes by the
short leap in all directions, forwards, laterally, and backwards. It is compul-
sory to take, under penalty of the huff. When a man reaches the opposite edge
of the board, it is promoted (name not given) ‘without receiving any distinc-
tive sign as a rule’ and moves any distance along any marked line through
the point it occupies and in any direction: nothing is said as to its power of
capture.
70 WAR-GAMES [CH. 4

Two triangles are also added to the quadruple alquerque board, each adding
six points as in the triangles of Fig. 28, and the two diagonals of every cell in
the quadruple alquerque are added to give the following game:
4.2.36. Simalur, an island off the west coast of Sumatra: Satoel (E. Jacob-
son, in Tijdschrift voor Indische Taal-, Land- en Volkenkunde, lviii, 1919, p. 80).
Each player has forty-six pieces arranged as in alquerque, leaving the central
point alone unoccupied.

A AR Eh FIG. 3Id. BOARDS FOR INDIAN WAR-GAMES

FIG. 31b. PRETWA

Games with the same rules of move and capture as in the alquerque games
are also played in India on other boards (Fig. 31a and 315), all the inter-
sections of lines except the central one being occupied by men (as in Figs. 27
and 31).
4.2.37. India, Lower Bengal: Lau kata kati (J. M. Datta, 167); United
Provinces, Karwi Subdivision: Kowwu dunki (E. de M. Humphries, 117).
Played on board of Fig. 31a, A, between sides of nine men.
4.2.38. India, Central Provinces: Dash-guti (H. C. D. Gupta, },
165); United Provinces, Karwi Subdivision: Kowwu dunki (E. de M.
Humphries, 117). Played on Fig. 314, B, between sides of ten men.
4.2.39] WAR-GAMES 71

4.2.39. India, Central Provinces: Egara guti (H. C. D. Gupta, c, 211).

Played on Fig. 31a, c, between sides of eleven men.
4.2.40. India, Behar: Pretwa (A. G. Shirreff). Played on Fig. 316, formed
of three concentric circles with six diameters, between sides of nine men
which are arranged on three consecutive diameters, leaving the centre of
the circles empty.
4.2.41. India, Central Provinces: Gol-skuish (H. C. D. Gupta, b, 165).
Played on a diagram of seven concentric circles with six diameters added.
Each player has twenty-one men arranged on three consecutive diameters.
4.2.42. Siam: Mak yep (Capt. Low, 380). ‘Mak yep is a game played with
fourteen counters on sixteen squares—one of the players must take off five
counters, so that he shall not leave one on the board—a situation to be taken,
for if one is so situated that it can go over the head of another to an empty
square, the first player loses.’ Except that the leap capture is used, I can make
nothing of this description.
A similar game to alquerque, but on a different board (see Fig. 32) is
played by some American Indian tribes. Mrs. M. C. Stevenson (who collected

FIG. 32. AWITHLAKNANNAI

the game) was told that the board had been introduced from Mexico where
it was used for a hunt-game, and that the Zufiian Indians had invented the
war-game.
4.2.43. American Indians, Zufiian stock, Zufii Indians, New Mexico:
Awithlaknannai, ‘stones kill’ (Culin, g, 801). It can be played .on two
boards, the awithlaknan mosona, with twenty-five points arranged in three
rows which are joined by lines so that three lines meet in each point, or on a
larger board, kolowis awithlaknannai (kolowis = serpent) which is double the
length, giving forty-nine points, arranged in the same way as the smaller
board. On the smaller board, each player has twelve men, and on the larger,
twenty-four, which are arranged so as to leave the middle point empty and
each player has one row entirely and the right-hand half of the middle row.
Men move one step along any line through the points on which they stand
and capture by the short leap.
72 WAR-GAMES [CH. 4

3. GAMES WITH THE LEAP CAPTURE AND DIAGONAL MOVES

4.3.1. Draughts. A game of European origin, played on the chessboard

and using cells of one colour only, which was invented in the twelfth century
A.D., and probably in the south of France.
(a) NOMENCLATURE. The dates of the earliest and, if obsolete, of the
forcat latest instances known to me.
1. The game. Latin: ludus dominarum (16th c.). French: dames (1507-c. 1750). Formerly
two varieties: jeu plaisant, plaisant (1570-1668) and jeu forcé, forcé (1535), Or
forcat (1542-1668). Earlier still, fierges (ante 1245).
Provengal: jeu de damo. Formerly forgat.
Catalan: dames. Formerly marro, marro de punta (16th c.).
Spanish, Portuguese: damas (1547). Formerly marro, marro de punta (1547-1650).
Earlier farisia (12th c.).
Basque: damen jokoa.
Italian: dama (1750). Formerly donne (1527), dame (1611). Dialectally in Sicily:
marrella (1617).
English: draughts (c. 1400). Formerly jew-de-dame (c. 1380), dames, dammes (1580-4).
In U.S.A. and dialectally in Eastern Counties and Isle of Wight, checkers.
Earlier, ferses (c. 1369).
Scotch: dams.
Dutch: damspel. Formerly dammen (1573), damen (1611).
German: damespiel. Formerly dammenspiel (1617), damenspiel (1655).
Danish, Norwegian: damspil.
Icelandic: damm, damspil (18th c.).
Swedish: damspel. Formerly damenspil (1668), damm (1743), dam (1745).
Finnish: damapelissa,
Czech, Hungarian, Slovene, Serbian, Rumanian: dama.
Polish: dame (1641).
Greek: ntama.
Turkish: dama (1694).
Russian: shashki (1684). In the Ukraine, damki.
Malay: dam blas.
2. The board. French: damier (1573).
Provengal: damié, damic.
Spanish: tablero (1590).
Italian: damiere (1597).
English: draughtboard (1681), draughtsboard. Formerly draught-table (1756).
Scotch: dambrod, damboard (1779).
Dutch: damberd (1571), dambord (c. 1682).
Danish: dambrat.
Icelandic: dammbord.
Swedish: dambrade.
Czech: damovnice.
Russian: shashechnitsa.
3. The draughtsman. French: pion (1668). Formerly and still dialectally, dame.
Provengal: damo, pioun.
4.3.1] W AR-GAMES 73
Spanish: peon (1590).
Italian: pedona, pedina. Formerly donna (1526), dama (1588).
English: man (1681). Also draught, draught(s)man. Dialectally, checker, checkerman.
Scotch: dam.
Dutch: dam (1627). Also damm (1571), damsteck (1627), dammeler (1671), damschijf.
German: stein (1617). Also damenstein (1744), scheibe, scheibichen (1744).
Danish: brikke, steen, dambrikke.
Swedish: bricka. Formerly dam (1734).
Polish: dam (1747), dama, kamien; ? biera, bierka.
Czech, Serbian: kamen, ‘stone’.
Hungarian: damko.
Russian: shashka.
4. The King. French: dame (1750). Formerly dame damée, damée (1611-1754), dame
couronnée (1727). Earlier roi (1243).
Provengal: damo damado, damat, damado, damé.
Spanish: dama (1590), peon coronado.
Italian: dama (1617). Formerly damata (1694), pedina damata (1837).
English: king (1681).
Dutch: dame.
German: dame (1627). Formerly dam (1694), doppelstein.
Danish: dambrikke.
Swedish: dambricka (1734), dam (1847), dame (1838).
Polish: dama (1641).
Czech, Hungarian, Serbian, Rumanian: dama.
Finnish: rouva ( = dame).
Russian: damka. Formerly doved.

(b) POSITION OF BOARD. This varies from country to country, and has
varied at different times in the same country.
1. A man stands on hr and hr is white: Spain, Italy in the eighteenth
century.
2. A man stands on hi and hr is black: Italy in the nineteenth century.
3. A man stands on ar and a1 is white: France, England in the eighteenth
century, Germany in the seventeenth century.
4. A man stands on ar and ar is black: Scotland, England in the nine-
teenth century, Holland, Germany since 1744, Denmark, Sweden, Poland,
Russia.
A preliminary examination of the nomenclature as set out above throws
light on some important features in the history of draughts: (1) There is no
evidence that draughts was played outside of France, England, and the
Spanish Marches before 1500; the first mention of the game, as played in
Italy, dates from 1527, and elsewhere in Europe is later than 1550 and implies
a spread eastwards from France. (2) The name of the draughtsman was bor-
rowed from the name of the queen in chess, at first fers, and when dame took
its place in chess, draughts followed suit, and the draughtsman changed to
dame also. In each case, draughts was named by the plural of the name of the
74 WAR-GAMES [CH. 4
chess queen. (3) The original name for the draughts king was in France roi, but
this passed out of use before 1500, and the king was known as dame damée, or
simply damée. The change of names, restricting dame to the king and using
pion for the man, was made in the second half of the seventeenth century in
France, and only became general there after 1750. In Spain the change had
been made by 1590.1 Except in Spain, dame before 1600, only means the man,
and never the king.
There are three references to a game of ferses, the first being in an Arabic
anthology of poems written by the Moors of Spain, the Kitab al-mutrib min
ash‘ar ahl al-Maghrib, compiled by the philologist, ibn Dihya (born 1149 in
Valencia, died 1235), of which there is a manuscript in the British Museum
(Or. 77). This work includes a list of the works of another Moorish author,
ibn Sharaf, among which is his work on games, with the game farista, which
means the player’s queen (malika), wherewith one plays as with the chess,
which work belongs to the most remarkable productions of the period.
The second reference occurs in the Chronique of Philip Mousket (1243),
which often makes use of metaphors drawn from games, and in the course
of a panegyric on King Philip Augustus (1190-1233), Mousket exclaims
(23617-20):
Cis n’estoit mie rois de gas,
Ne rois de fierges, ne d’escas (chess),
Ains iert 4 droit fins rois entirs,
Rubins, esmeraude et safirs.

The third reference occurs in Chaucer’s Book of the Duchess (c. 1369),
lamenting the death of John of Gaunt’s wife, Blaunche, Duchess of Lancaster.
In this poem (lines 617-748), Chaucer relates a dream in which he met a
knight who bemoaned that false Fortune was playing chess with him and
had taken his fers (queen), and when he saw his fers taken, had exclaimed:
Alas! I couthe ne lenger pleye,
But seyde, farwel, swete, y-wis,
And farwel al that ever ther is.

Chaucer then tries to comfort him and tells him that his grief would have
been exaggerated
Thogh ye had lost the ferses twelve,
and ends by declaring
But ther is noon a-lyve here
Wolde for a fers maken this wo!

Now, while the knight’s parable is entirely consistent with chess, Chaucer’s
reply is not. By the loss of the ferses twelve, he can only mean the loss of
? We know nothing about the Spanish nomenclature until the middle of the sixteenth century
when the man was peon (possibly because Spanish players used pawn-shaped men as late as 1684),
and the king dama.
4.3.1] WAR-GAMES 75
everything. Skeat, in his note on the passage, realized this, and interpreted
ferses twelve as meaning all the chessmen except the king, but this explanation
is too artificial to carry conviction for an instant. The only satisfactory
explanation is that Chaucer and the knight were playing at cross-purposes,
and that while the knight was thinking of chess, Chaucer was thinking of
the game of ferses. The knight’s reply, in effect, ‘you don’t know what you
are talking about’, gains force by this explanation.
Each of these references contributes something of importance to our
knowledge of the game of ferses. From ibn Dihya we learn that the game
was neither of Moorish nor of Spanish origin; the word farista is an arabicized
form of the Provengal fersa, and was so named because the men with which
it was played were ferses or chess queens. From Mousket we learn that
promotion was possible, and that by promotion the fers became a king. It
follows that the men were limited to the forward diagonal moves of one
step. From Chaucer we learn that the game was a battle-game, in which
each player had twelve men. Only the method of capture is undescribed.
The conclusion that ferses was draughts seems inevitable, and the life of the
name ferses, ¢ 1150-1400, confirms this conclusion. What then were its
parents?
Undoubtedly one parent was chess which provided the board and the
name of the draughtsmen, fers as long as it was used in chess, then its successor
in French, dame, and after the death of the medieval chess, peon in Spanish,
pion in French, and pedina in Italian. The other parent seems to have been
alquerque, which provided the number of men and the method of capture.
I have already (p. 7) shown that the chequered chessboard invited the
transfer of other games on lined boards to the cells of one colour on the
chessboard. This view is supported by the use of marella for draughts in
Sicily (Carrera, 36) and of marro and marro de punta in Catalonia and Spain,
which survived in vulgar use as late as 1650 (J. G. Canalejas, Libro del Iuego de
las Damas, Garagoga 1650, iia).
In this primitive draughts, as played in France and England, there was no
compulsion to take when taking was possible. This simple form survived in
the provinces of France and in the rural counties of England into the second
half of the seventeenth century. It was known in France as le jeu plaisant de
dames or briefly as plaisant in contrast with the jeu forcé or forgat in which
capture was compulsory. Mallet (p. 324) describes plaisant as ‘un jeu d’Anfant
qui ne mérite point de Loix’, and Hyde (p. 187), when explaining the method
of capture, advises players to refrain from taking, if it involved the loss of
a man.
Five mentions of draughts between 1200 and 1500! do not point to any

1 The three quoted above and two more in English versions of French works (where the
original French versions do not mention draughts): Destruction of Troy, c. 1400 (EETS.), line
1622: ‘The draghtes, the dyse, and oper dregh gaumes’; and Sir Ferumbras, c. 1380 (EETS.),
line 2225: ‘iew-de-dame’.
76 WAR-GAMES [CH. 4
great popularity in the Middle Ages. There are many ecclesiastical ordinances
and town custumals of the period, but none mentions draughts. It is difficult
to resist the conclusion that the game cannot have been very widely known
before 1500.
4.3.2. England: Draughts; Scotland: dams; France: dames. This only
differs from the medieval game in one particular. Capture is compulsory,
under penalty of the huff, which was introduced in the early sixteenth century
and raised draughts to a parity with chess. As Eloi d’Amerval says in his
Grande Diablerie, 1507, U1. xiii:
Comme au jeu d’eschecz ou de dames
Qui sont beaulx jeux non pas infames.

Rabelais, in the list of Gargantua’s games (I. xxii, written ante 1535), includes
both dames and forcé (in later editions forgat), and in a later book (tv. xi, first
printed in 1547) he uses the verb damer, ‘to crown a man’, in the sense of ‘to
cap a story —‘je dameray cest cy.’ Cotgrave (1611) defines forgat as ‘a game
of draughts, wherin one must take his aduersarie when he may, or else he
himselfe is taken’. The forsat game is the English game of draughts.
The introduction of the compulsory capture in draughts was made in
France, and seems to have synchronized with the changes in chess which
created our modern chess. In the course of the sixteenth century the reformed
draughts spread rapidly from France to Italy, England, the Low Countries,
and Germany, and may have led in Spain to the change of the name of the
game from marro to damas. Le donne is mentioned in a list of Italian games in
1526 (Fiske, 187) though Cardan does not seem to have known the game. In
England, some Elizabethan authors use dames or dammes, but draughts was
strong enough to withstand the challenge. In the Low Countries, we meet
with dambert in a Flemish lampoon on the Duke of Alva, which refers to his
coming to Antwerp in 1568:
"t Antwerpen is hy ghecomen,
Sijn dambert heeft hy mede ghebrocht

which refers, not to the draughtsboard, but to Alva’s coat of arms (a shield
checky), which still adorns the gateway of the citadel that he built to over-
awe the town.
The life of the reformed draughts in France was only brief. Mallet (p. 300)
says it was much played there by mathematicians, but it had to meet the
challenge of a variety, invented in Paris about 1650, named grand forcat to
which Mallet devoted the first French work on the game. In this game the
idea of compulsion was carried farther: the player must take when he can or
immediately lose the game, and if he can take in two directions, he must
take in the direction in which he can make the greater number of captures,
and if the number is equal in both directions, then in that in which he takes
the greatest force, a rule known later in Italy as il pik col pit. This game,
A382] WAR-GAMES hi

however, failed to secure acceptance outside of Paris, and even in Paris was
obsolete before 1700. The second French work by Quercitano! (1727) knew
nothing of grand forcat, but tried to kill a new game which had appeared at
the Hotel de Soissons, by calling it ‘un pur badinage de l’orgueil et de l’esprit
humain’. In less than twenty years this new game, now known as Polish
draughts, has ousted the older game from the cafés of Paris and Holland.
The forgat game ran a more favourable course in England where it came
into prominence with the Restoration (1660). Lucas names two of the leading
players before 1700: Col. Panton (died 1681) and Jonathan Laud (died 1704).
The latter was ‘such a great artist at this game, that he would give any man
a guinea that would bring a person to play with him for three guineas, till at
last he was so noted for his ingenuity at Draughts, that none would play
with him’.
Draughts is briefly described by Randle Holme (a, 67), but William Payne’s
Introduction to the Game of Draughts, 1756, was the first English work to deal
analytically with the game, and marks the beginning of the great modern
popularity of draughts in England, Scotland, and the U.S.A.
The Scotch terminology of draughts shows that the immediate source of
the game in Scotland was Holland, and points to an introduction in the
seventeenth century as a result of the close intercourse between the Low
Countries and Scotland during the Thirty Years War, when so many Scots
served abroad in the Protestant armies.
Draughts also spread from the Low Countries to Germany where the first
mention of Dammen-Spiel was made by Selenus (1616, p. 32). The earlier
German draughts was identical with the English game, but since 1800 other
varieties have obtained a footing in Germany without wholly ousting the
English game.
The Scandinavian countries, Poland and Bohemia, received their draughts
from Germany in the course of the Thirty Years War. The earliest mention
of the game in Sweden is in a work of 1640-7 and in Poland in 1641. The
game only reached Iceland in the middle of the eighteenth century, and Isl.
Gatur (p. 320) says that itis still not widely played. Danish and Swedish manuals
of the early nineteenth century describe the English game as the ruling
variety, but include accounts of the Spanish game and of an intermediate
type which seems to have been invented in Denmark or Sweden.
The English game was carried by earlier settlers to North America, where
it is known as checkers (which is also used in those parts of England from
which many of the Pilgrim Fathers emigrated). The French settlers in
Quebec call it le jeu franc. The game has spread to some North American
Indias (Culin, g, 792, 797).
1. Micmac stock, Nova Scotia: Adenagank, ‘moves’. The men are either
smooth disks or square.
™ His real name was Mars; he was an official in the Pay Office of the Artillery who was known
to contemporary draughts-players as le Maftre de Tout.
78 WAR-GAMES [CH. 4
2. Wabanaki and Passamaquoddy stocks, Maine, New England: name
not recorded. Mrs. W. W. Brown (‘Some indoor and outdoor Games of
the Wabanaki Indians’ in Trans. Royal Soc., Canada, 1888, p. 41) says that
nearly all the Wabanaki Indians are more or less proficient at checkers or
draughts.
3. Siouan stock, Omaha, Nebraska: Wakanpamungthae, ‘gambling
bowed head’. English draughts; learnt about 1873.
4-3-3. Italy: Dama; in 1526 le donne and in 1580 dame, both plural. A
diagram of the board with the arrangement of the men (men on ar, &c.)
occurs in a Latin manuscript (Perugia, Communal Library, 1. 27, f. 1634, of
the end of the sixteenth century, with the title ‘Ludus dominarum D(ifficilis)’.
The oldest description of the game is given by Mallet (ch. xviii). It only
differs from the English game in two particulars: the man cannot take a king,
and captures are governed by the rule il pin col pit (p. 76).1 Italian sailors
carried the game to the Levant, hence the modern Greek ntama and the
Turkish dama.
4.3.4. Spain: Damas. In this game the man retains the move and capture
of the English game, but the king is given an unlimited diagonal move, that
of the bishop in modern chess, and in making a capture the king can be
placed on any cell beyond the captured man, provided all the cells between
this cell and the one occupied by the captured man are vacant. It is com-
pulsory to take the maximum number of pieces possible by the move. The
capturing king must complete its move and be quitted before any of the cap-
tured pieces is removed, and a taken, but unremoved, piece exercises the
same restraint on move as an uncaptured piece.
This modification appeared first in Spain and already existed before the
publication of the first Spanish work on draughts (Antonio Torquemada, El
Ingenio, 0 Juego de Marro de Punta 0 Damas, Valencia 1547). It is tempting to
associate the change in the king’s move with the adoption in chess of the
modern move of the queen which was made a little before 1500. There is
no evidence that Spanish chess-players had called the medieval queen dama
before the introduction of the reformed chess: Lucena (c. 1497) still uses
alferza in some of his problems of the older chess.
Torquemada’s work is now lost, but from its title it was a collection of
problems, not of games. It was followed by four other works between 1591
and 1684, which point to a higher standard of play than was reached in any
other country. In all these works, the man is peon and the king dama. The
first of these works (Montero, 1591) gives some attention to games in which
one player concedes the odds of a man, and the next three, Valls (1597),
Canalejas (1650), and Garcez (1684) all include ‘games with made kings’ in
which each player replaces one or two of his men by kings; later works omit
this variety.
' Giuocatore in Conversazione, Milan, 2nd ed. 1820, says that this rule was not observed by
players in Milan.
4.3.4] WAR-GAMES 79
Among the varieties of draughts that are now played in Germany is one
with the Spanish moves of man and king. This is mentioned in J. B. Montag’s
Vollsténdiger Unterricht im Damen-Brettspiel, c. 1850 as usual in many parts
of
Germany, and Brockhaus’s Conversations-Lexicon (14e Aufl. 1901), gives it as
the main German variety. It had already spread to Denmark a half-century
earlier. A Swedish manual of 1847 says that it is a matter for agreement
when playing dam whether the king is to have the short or long king’s move.
The Spanish game is also played in the Philippines (Culin, f, 648) under
the name dama, but on the lined board of Fig. 33. This simplification was
first suggested by J. G. Lallement in 1802.
Spanish draughts is also played in Hawaii
where it is called moo and played on the
chessboard (Culin, c, 244).
4.3.5. Denmark: Makvar; Sweden:
Marquere. This is a modification of the
Spanish game in which the man has
the power of taking at a distance like the
Spanish king, but in the forward direc-
tions only (S. A. Jérgensen, Nyeste Dansk
Spillebog, 1802; Hand-Bibliothek for Salls-
kapenojen, 1838).
4.3.6. Poland: Dama, formerly (1641) FIG. 33. DAMA (Philippines)
dame (pl.); Russia: Shashki; Ukraine:
Damki. In these games the king has the Spanish king’s move, and the man
is allowed to take in both the forward and backward directions. In this
they resemble Polish draughts on the 10x10 board, and English writers
(who name this variety ‘minor Polish’) have generally assumed that it
originated in the transfer of Polish draughts to the chessboard. This, I believe,
is a mistake; we know that draughts was played in Poland as early as 1641
and in Russia at the court of Peter the Great’s father, and there is some evi-
dence that the move of the man in Polish draughts was suggested by a Pole
resident in Paris (see below).
The Polish and Russian games on the 8 x 8 board differ in one particular.
In the course of a consecutive series of captures, it may happen that the man
reaches the crownhead (the opposite edge of the board) and still has captures
open to it. In the Polish variety, the man only becomes a king when it comes
to rest on the crownhead; if further captures are possible, he must make
them as a man. In the Russian variety, the man is promoted king directly it
reaches the crownhead and, if further captures are possible, it makes them as
a king. This is the rule in the first Russian work on draughts, Petroft’s Ruko-
vodsko, 1827, p. 4.
Draughts has spread to many of the native tribes in Siberia, and is played
by the Samoyedes, Tungusians, and Yakuts, but none of the mentions that
I have seen gives any description of the games.
80 WAR-GAMES [CH. 4
As a result of the Partitions of Poland in 1772, 1793, and 1797, parts of
Poland were incorporated in Prussia and the Polish variety on the chessboard
became domiciled in Prussia. It is described by J. B. Montag and inJ. Dufresne,
Kleines Lehrbuch des Damespiels, 1884.
4.3.7. Polish draughts. The enlarged form of draughts between sides of
twenty men on a board of 10X10 cells was developed about 1725 and was
first played in the cafés of Paris in 1727. The first mention of the game by
name occurs in the London periodical, the Craftsman, on 15 September 1733,
which promised an early article on Polish draughts ‘where you will see the
whole board engaged in the important business of making Kings. There you
may observe the whole Art of Intrigue and Bribery, Fraud and Force. This is a
game of skill, but more confused and irregular than that of Chess.’ The
promise was never fulfilled. The game caught on, and soon displaced the
game on the chessboard in France and Holland. Today it is the draughts of
France, Belgium, Holland, and the French-speaking cantons of Switzerland.
As so often has happened in the case of a new game, no contemporary
writer recorded its origin, and when Charles de la Condamine asked for
information in a letter to the Mercure de France, July 1770, only vague and
unsatisfactory replies were forthcoming. These are reproduced in Manoury.
The only definite statement on the matter was that K. Stanislas Leczinski,
who lived in France after his deposition, had told de la Condamine that the
game was known in Poland as le jeu de Dames frangoises, which certainly
implies that the game had been invented in France.
The most circumstantial of the tales recorded by Manoury is that the
game was invented at the court of the Regent, Philippe II of Orleans (1713-
~23), by an officer of his court in collaboration with a Pole who went by the
name of the Polonese. The latter suggested that better ‘strokes’ would be
possible in draughts if the man could take backwards (as presumably it could
in his native Polish variety), and the former that still better strokes would be
possible on a larger board. This may well have been the way in which the
new game was evolved.
The existing standardized Polish draughts gives the man a move of one
cell in the forward directions and the man can take in both forward and
backward directions by the short leap. The king has the long move and
capture of the king in the Spanish variety. If a player has a choice of captures,
he must choose the one in which the greatest number of captures is made.
Captured men are only removed after the completion of the move, as in
Spain. The man is only promoted when it comes to rest on the crownhead;
if on reaching the crownhead, other captures are possible qua man, these
must be made and the man remains a man until he again reaches the crown-
head and no further men are en prise.
These rules have varied in the past, and still do in some places. Manoury
(1787) laid down that captures must be il pitt col pit, but this was dropped in
France before 1811, though it was still in force in parts of Belgium in 1886.
4-3-7] WAR-GAMES 81

In parts of Holland, if an equal number of captures can be made by king or

man, the king must take. Around Ghent, the man could not take backwards
(Gazette du Jeu de Dames, 1886-7, p. 184). Van Embden (1785) said that in
Amsterdam, if a man in a series of captures reached the crownhead, he became
a king and the move ended although other captures might still be possible;
this rule, however, was abandoned before 1850.
The Dutch carried Polish draughts to their East Indian possessions and the
game has been adopted by the Malays of Malaya and Indonesia under the
name of dam blas, ‘huff draughts’. A careful account of this game by H. O.
Robinson appeared in the Cheltenham Examiner, 19, 28 Nov. 1903. The man
cannot take backwards, and when there is a choice of captures, the one which
takes the greater number of pieces must be chosen.
The invention of Polish draughts led to a considerable experiment in Paris
of which accounts are given by Manoury and by Twiss (Misc., ii. 172-5). Only
one of these had any vitality, Babylonian, which about 1736 caused Polish
draughts to be neglected for some time, and was still played in Friesland,
Holland, as late as 1886 (Gazette du Jeu de Dames, 1886-7, p. 184). It was
played on the 10 10 board (a1 white and a white man on ar) using only the
white cells, with sides of twenty men which moved like the modern queen
in chess (moves to black cells, of course, excluded).
4.3-8. Canadian draughts. In the French-speaking provinces of Canada,
draughts is playedintwo ways, le jeu franc (the original French forgat) and lejeu
de dames canadien, played on a board of 12 x 12 cells between sides of thirty
men. The moves of man and king, and the rules of capture are those of Polish
draughts, but the huff (asin Holland) has been abolished since 1880, followingan
unfortunate dispute in the course of a match for the championship of Canada.
It is not known when this game was first played in Canada, but a column
dealing with it was begun in L’Opinion publique, 2 Dec. 1875. The idea of an
enlarged Polish draughts is, however, older. Twiss (Misc., ii. 175) says that
boards of 12X12 cells were on sale in London in 1805. A game on this
board is said to have been popular with British soldiers stationed in India,
and is played both in south India and in the coastal parts of Ceylon, under
the name dam. H. O. Robinson told me that he had seen dam blas played on
the 12 x 12 board in Penang, about 1896.
4.3.9. India, Ceylon: Dam (Parker, 584-5). Played onachequered board
of 144 cells with sides of thirty men which move one step diagonally for-
wards and backwards, and take as in Polish draughts. On reaching the
crownhead, a man becomes a king with the king’s move of Polish draughts.
There are two minor varieties of draughts which are occasionally played.
4.3-10. Britain: The losing game; France: Coc-Imbert (Mallet, 438);
Germany: Das verkehrtes Damenspiel (F.T.V., Das erklarte Damenspiel,
1744; ch. ix); Russia: Poddavki, ‘give away’ (A. D. Petroff, 1827).
Played on the 8x 8 board with twelve men against twelve or one, and on
the 10x 10 board with twenty men against one. One player with the larger
6356 G
82 WAR-GAMES (CH. 4
number of men endeavours to force his opponent to capture all his men.
Cf. the variety of chess called forgado (Alf. 5a) and the variety of mancala,
lontu-holo (7.3.5), played by Bush negroes in Dutch Guiana.
Coquimbert has been used in French for other losing games. Thus
Rabelais includes among Gargantua’s game one called coquimbert which
Florio (1598) identified as the Italian carica l’asino and Cotgrave (1611) as
the English losing loadum, a card-game.
4.3.11. Britain: Diagonal draughts. Played on the 8 x 8 board with two
sides of twelve men which are arranged so as to leave the double diagonal
clear (J. A. Kear, Sturges’ Guide, 1899, xviii).
4.3-12. Siberia, Kalmucks: Mingma (P. S. Pallas, i. 157). ‘They have also
a kind of draughts (mingma) in which men are placed on the white squares
and the black are left bare.’

4. GAMES WITH THE LEAP CAPTURE AND ORTHOGONAL MOVES

4.4.1. Turkey, Palestine, Egypt: Atlanbaj, ‘leaping game’; now called

Dama (Hyde, 180; MS. Berlin, Landberg 806; modern Arabic manuals).
Played by two persons on the unchequered 8 x 8 board, each having sixteen
men which are arranged on his second and third rows. The man moves one
cell in the forward or lateral directions, and on reaching the eighth row
becomes a king (Hyde, padishah) and obtains the rook’s move in chess. The
man captures by the short leap in the forward and lateral directions, and the
capture may be multiple. The king takes by the long leap and may be placed
on any of the succession of empty cells beyond the captured piece to make
further captures. Captured pieces are removed as taken, and the player is
obliged to take the maximum number of pieces en prise, or loses the game
forthwith (since captures are removed as made, it is difficult to see how this
rule can be enforced, when as many as twelve men may be en prise, and it is
very easy to lose one’s way in playing the stroke). The game is won by the
first player who takes or immobilizes all his opponent’s pieces, or reduces
the forces to king v. single man. I owe this rule to Hilmi Samara.
4.4.2. Japan: Tobi-shnogi, ‘jumping chess’ (Cho-Yo, 194; Prof. Tsuboi).
Two persons play, each having twenty men which are placed on the points
of the two back rows of the Japanese chessboard of 9 x 9 cells. The Japanese
use their chessmen as men for this game, giving them all the same power of
move. Men move one step forward and take by the short leap. A children’s
game.
4.4.3. Egypt: Siga (Parker, 604-5). Played by two persons on a board of
5x5 cells or five rows of five holes marked out on the ground. Each player
has twelve men which are entered, two at a time, in alternate turns of play,
until all the cells except the central one are_filled. When all are entered, the
game proceeds by alternate moves of a single man, one step in any of the
four orthogonal directions. Men capture by the orthogonal short leap which
may be multiple. The game appears to be a modification of Siga (4.1.7).
4.4.4] WAR-GAMES 83

4.4.4. Somaliland, Issaq tribe: Korubéddo, ‘high leap’; Darod tribe:

Lorkabéd, ‘jump over’ (G. Marin, a, 506). Played in the same way as 4.4.3,
but the player who enters the last two men begins the subsequent play, and
captures are not compulsory.
4.4.5. Mandinko and Fula tribes, Gambia Valley, West Africa: Choko
(Parker, 604). Two persons play. The board of 5x5 holes is set out on the
ground. Each player has twelve men, one using sticks about 5 in. long (kala),
the other using sticks about 3 in. long (bond). The men are entered one at a
time in alternate turns of play and it is not necessary for all to be entered before
moves can be made, but whenever one player enters a man, his opponent
must do the same. Moves usually begin before all the sticks are entered. A
man can move one step in the orthogonal directions and takes by the ortho-
gonal short leap. Only one man can be taken in a move, but the player who
makes the capture then removes any second man at choice.
I think this must be the game which Borel saw (Avelot, b, 11, quoting
Bérenger-Feraud, Les Peuplades de la Sénégambie, 1879, p. 77).
4.4.6. Libyan Desert: Helga (C. D. Belgrave). This game, which Bel-
grave describes as ‘rather like draughts, played with
onions and camel-dung on a board which is marked
out in the sand of the desert’, is probably the same
as 44.5.
4.4.7. Liberia: Queah tribe; name unrecorded
(J. Biittikofer, IAFE., i. 91 and Fig. 16). Two
persons play on a board made of interwoven twigs
to form a board of thirteen holes. Each player has
ten sticks, one side having the tops cut aslant for rig. 34. BOARD FROM
the ‘men’, the other having the tops cut straight for LIBERIA
the ‘women’. Each player enters four of his pieces,
as shown in Fig. 34. The pieces move one step diagonally and take one
enemy piece at a time by the short leap, and his opponent then enters a
man to restore his number of four pieces on the board until his stock is
exhausted.

5. GAMES WITH THE REPLACEMENT CAPTURE

4.5.1. Chess and its derived games. Chess, the most elaborate and charac-
teristic of all battle-games, was invented in north-west India after the close
of the Hun Domination, probably round about a.p. $70.
As I have already dealt fully with these games in my History of Chess,
I shall confine myself to one aspect only of chess which may throw light
on the invention and history of other board-games, the diffusion of chess to
other countries and the speed with which this took place. This diffusion took
place in the following directions.
1. Westwards to Persia before 600 and to the Muslims by 650 as a result
of the conquest of Persia. The Arabs then became the great disseminators of
84 WAR-GAMES [CH. 4
chess, carrying it to Egypt by 720, Morocco and Spain by 750. Christian
Europeans were playing chess in the Spanish Marches by 1010, in Italy and
south Germany by 1050, in France before 1070, and in England before rr1oo.
It reached Iceland before 1230. The Byzantine Empire knew chess by 800,
and Russia obtained the game from Constantinople, or possibly direct from
Persia in the course of trade in the twelfth century. Muslim invasions of India
carried the Muslim chess back to India before 1ooo and Persian refugees
carried their chess to Bombay. N. Nigeria, Abyssinia, Zanzibar, and Mada-
gascar all obtained their knowledge of chess from the Arabs.
2. North-west by Kashmir to central Asia and thence north to the Mon-
gols and other Siberian peoples, and east to China which chess reached about
750. From China it travelled north to Korea and Japan about 1000, and south
to Siam, Annam, and Malaya.
3. North to Tibet and south to Deccan and Ceylon, and thence to the
Malay races.
4. East to Burma, Siam, and Annam.
4.5.2. France, England, Italy, Germany: Rithmomachy (Lat. rithmo-
machia), ‘battle of numbers’ or Ludus philosophorum, ‘the Philosophers’
Game’. An arithmetical game, invented in the twelfth century, probably in
France, and still played in the early seventeenth century when Robert Burton
(Anatomy of Melancholy, 1621, m1. i. iv) included it in a list of games played in
England during the winter. It was much belauded in the Middle Ages, but,
as it involved some knowledge of the Progressions, it was probably confined
to the more learned classes. It is described in the Vetula (1. xxxv) and was
the subject of a numberof books: 1482, John Shirwood (Bishop Durham,
1483-93) who wrote his book at the instance of his patron, George Neville,
Archbishop York (1465-72); 1426-1512, Jacobus Faber Stapulensis; 1554,
1556, Claude de Boissiére (Buxerius); 1563, William Fulke, who translated
de Boissiére into English; 1572, Francesco Barozzi; 1616, Gustavus Selenus,
who translated Barozzi into German; and 1705, R. A. V. W. Gorlitz.
Played by two persons on a double chessboard 8 x 16 cells. The players sit
facing the longer sides of the board although the pieces are arranged along
the shorter rows. Each player has twenty-four men, known as White (or
even) and Black (or odd).
Eight men on each side are round disks, eight are triangles, and eight
squares, all coloured white on one face and black on the other so that they
can be used by either player. Every man also bears a number which is
repeated on the reverse face. These numbers are chosen on arithmetical
principles, and the explanation of these numbers is worked out at length in
the books on the game. The white are based on the sequence 2, 4, 6, 8 and
the black on the sequence 3, 5, 7, 9. The numbers on each man are shown
in Fig. 35.
On both sides, one of the squares (91 on the white side and 190 on the
black) is replaced by a pyramid or pile of men of the different types. Thus
 gh ALAS +
”
pe Ge aed
S eas a <
8 ©3448
ad
am3
8 "tee
sO 2.8 8S©
G ® og ang
eas Sg ‘Su LL Yj} Z Z , | peyle
| Ys Aa ep
fa egetSg Wy Y y y mer
Yi} 8B sueuo Li au 4. 4. a
ee
pee
eS
es eh soeYj,
YY
oe
2.358
££55 aS SO fy A GK} A,

3a sa ()
7S re} SEY ate} leah .
ooo 2p .oe Y
17
eee
UW)
AO
om
a
a
ws G's
4 a, es y
aug aea gO W; f
SEERE RB
| “A aE Pe bad
F 200
oA
Pre
pe
es

8 = oo§$ 8 Uy
Wi
Oy
On
Ss
5 a CASisloge,
Ono
4
0-5 Go 4 “Y
“eh
&2v
oR SO OF ©
ges gab WY
ek Biko
ran (ls
Yer 1, u>13 Yi}
Tale ome
eee Y t= oe
Yalen fel Cy
a hae pe
aoa
mo
e ie detest
ys
Bo SB
.& % $68 oo
86 WAR-GAMES [CH. 4
(a) THE OLDER GAME (Shirwood). The men are arranged as in Fig. 35,
though it is open to the players to adopt a different arrangement. The round
moves one step to any adjacent empty cell. The triangle moves two steps
orthogonally or diagonally (but not like the knight in chess), the cell passed
over being empty. The square moves three steps orthogonally or diagonally,
both cells passed over being empty. The pyramid can move like any of the
layers of which it is composed. No piece can leap over any other piece.
In making captures, the number associated with the man comes into play.
Shirwood describes four ways in which captures can be made, by equality,
addition, multiplication, and blockade. The first three are connected with a
move, the fourth not.
1. Equality. A man, numbered n can make a legal move to a cell occupied
by an enemy man with the same number n. The enemy man is taken and
the capturing man takes its place. Since the only numbers that occur on men
of both sides are the square numbers from 9 to 81, and in every case the men
are of different shapes, no men can be mutually en prise, and the opportunities
for capture by equality are few.
2. Addition. Two or more men can both be played by a legal move to a
cell occupied by an enemy man, and the sum of the numbers on these men
equals the number on the enemy man. The enemy man is taken and one of
the attacking men takes its place, the player who takes being free to choose
which shall do so.
3. Multiplication. A man numbered n is x cells distant along an orthogonal
or diagonal direction from an enemy man with number xn and no man
intervenes between them. The enemy man is taken and the man with number
n takes its place.
4. Blockade. Every cell, to which an enemy man could move if it were
empty, is occupied by men, none of which is en prise to the blockaded man.
The enemy man is removed from the board and the player then makes a
move. Blockade is the only way in which men numbered 2, 3, 4, 5, 6, 7, 153,
and 190 can be taken.
Boissiére says that captured men are turned over, retained in hand, and
when the player thinks fit, are entered on the player’s back row and become
part of his army. Shirwood says nothing about this, except that the player
can use a captured man for the construction of the triumphs.
The rules of capture apply also to the pyramids, but in their case the layers
can be taken separately by an enemy man with the layer number, or the
whole pyramid may be taken by a man with the pyramidal number (the
total of the layers).
The prime aim of the game is the complete capture of the enemy pyramid.
When this has been accomplished, the player who has taken his opponent’s
pyramid proceeds to the triumphs. These are three in number, and they have
to be obtained in order. The first is to arrange three men in a row so that
their numbers form a series in arithmetical, geometrical, or harmonical
4.5.2] WAR-GAMES 87

progression. The second triumph is to arrange four numbers so that they

include two sets of three in different progressions. The third and greatest
triumph is to arrange four numbers so that they include sets of three in all
three progressions.
(b) THE LATER GAME (Selenus). The arrangement differs from Fig. 35, the
sides being arranged on the first four rows instead of on the third to the sixth
rows. The round moves one step forward (like the chess pawn); the triangle
two steps diagonally, but the intervening cell must be empty; the square
three steps orthogonally and diagonally, the intervening cells being empty;
the pyramid as any of the layers composing it when it makes the move, and
once in the game to avoid capture as the knight in chess.
Two other forms of capture are added to those given by Shirwood:
2. Subtraction. Two men can be played by a legal move to a cell occupied
by an enemy man, and the difference of their numbers equals the number of
the enemy man. The enemy man is taken.
3. Division. A man numbered n is x cells distant along an orthogonal or
diagonal direction from an enemy man with number nx. The enemy man
is taken.
A turn of play consists of either a move or a capture, but not of both
together. If a capture is possible, the player must make it, but the man that
makes the capture remains on the cell from which it makes the capture. Ifa
player omits to make a possible capture and makes a move, his opponent
huffs the man which ought to have made the capture and then moves. If
he fails to make the huff, the first player may then make the capture and
also make a move.
A pyramid may be taken as a whole or layer by layer. If a layer is taken
and the player with the pyramid has another man with the same number as
the layer taken, he may substitute it-for the captured layer. When a whole
pyramid is taken and the opponent turns it over to add it to his forces, it has
only the value of its lowest layer, thus the inverted pyramid 91 has the value
36, and the inverted pyramid 190 has the value 64. When a player has taken
his opponent’s pyramid and proceeds to play for a triumph, his opponent
still plays in his turn.
The game may be shortened in various ways; thus the players may agree
that the first to make a given number of captures, or to take a given total value
of men, or to take a given number of men with a given total value, wins.
6. GAMES WITH OTHER FORMS OF CAPTURE

4.6.1. Madagascar; Hova and Imerina tribes: Fanorona (Montgomery,

148-56); Betsileo tribe (Dubois). This game was developed from alquerque
! Faber and Boissiére have different powers of move though keeping to the number of steps
allocated to each type of man. They limit all moves to the orthogonal directions. Boissiére also
adds another form of move: round, one step diagonally; triangle, two steps orthogonally; pyra-
mid, three steps orthogonally or diagonally, and allows every man except the pyramid to move
once as the chess knight but no capture can be made on this leap.
88 WAR-GAMES [CH. 4
(Fig. 27) about 1680 by doubling the board and increasing the number of
men on each side to twenty-two and by changing the method of capture.
It played an interesting part in the rituals and Mrs. Danielli tells me that,
during the storming of the capital by the French in 1895, the Queen and

FIG. 36. FANORONA

people relied far more on the outcome of the official game which was being
played by the ritual professionals for victory, than they did on their armed
forces.
Played by two persons. The board is sometimes inscribed on a flat stone
and sometimes marked out on the ground. The men are arranged on the
points as in Fig. 36, the central point (foibeny, ‘navel’; cf. the Malay pusat on
the alquerque board) alone being empty. The game is played on the marked
lines and a man can move one step along any marked line through the point
on which it stands to a neighbouring empty point.
Ifa move ends on a point and the following point or points in unbroken
sequence along the line of movement are occupied by enemy men, or if a
man which has an enemy man on the next point on a line through its position
or an unbroken sequence of men along this line, moves away from this man
or sequence of men along the same marked line, the man or sequence of men
is taken by approach or withdrawal. Captures are compulsory, but on the first
move by each player only one such sequence can be taken. For the second
and later moves, a player can repeat these manceuvres and make further
captures in the same way by moves along another marked line. If a move
puts enemy men in both directions en prise, the player can only take in one
of these directions. “Don’t eat at both ends like a leech’ is a Malagasy proverb.
Each successive approach or withdrawal must be made along a different
marked line.
 4.6.1] WAR-GAMES 89
Montgomery gives the following game as an average specimen of native
skill. (I use : for ‘takes’ and give the number of men taken by each move.)
White Black White Black White Black
T-€2-e7:2 f4-¢523 14-1372) a3—b2 17 f2-g2 c2-d2
2 hi-hs 3 9-2 5 £3-<3 ba-brzr 18 f4-e3 21 - “b4ec5
f4-g3 21 6 e3-d4 b4-a3 19 €3-d4:1 a4-a3
g3-f2 : 1 97 d4-d3:1 a3-b2 Xe alice sit @biaby oi
f2—e2 : 4 8 g4-f4 bi-c1 21 g2-93 a3-b3
e2-€1 :1 9 i2-h2 bs-b4 22 93-3 b3-c3
e1-f1: 4 10 f4-e5 b4-a3 23 3-93 c3-d4
3 g2-g3:2 11 h2-g2 as—bs 24 g3-h4 d4-€3
g3—h4 : 1 12 g2-f2 b2-co 25 h4-h3 e3-f4
h4a-g4:1 ft-er:1 13 i3-h4 a3-b3 26 h3-i3 d2-€3
4 d3-€3 :1 14.d3-c3:1 ca4-c§:i 27 13=h3 e3-f2
e3--f2 : 2 c§-ds:1 28 h3-13 2-14
f2-93 21 15 h4-g3 c1-di 29 13-12 £4-93
g3-f4 : 1 16 g3-£4 bs—b4 and wins

When one player is defeated, the next game is played differently, the new
form of play being known as vela. The defeated player begins and the winner
proceeds to sacrifice man after man until he has parted with seventeen men.
During this play the winner refrains from making any capture and his
opponent may only take one man each move. Montgomery gives this
example of a vela game.
White Black White Black White Black
t d3-e3:1 c3-d3 14 d4-e4:1 g5-f4 20 e3-d3 :1
2 e3-f3:1 dé4-c3 15 e4-d4:1 g4-f4 dada 5
3 d2-e3:1 gs5-f4 16 d4-e4:1 is—hs d2-c3 : 1
4 e3-d4:1 hs-gs 17 f3-f4:1 = hg-gg: 1 c3—-b2 : 1
§ e2-e3 71 b5-c5 25-24 33 b2-a3 : 1
6.e1-d2.- 0 = 15ehs 18 i2-13: 1 h3—ha>2 a3—a4 21
7 b3-c3:1 c5—bs h4-g3 : I a4—b4: 1
8 ba—b3:1 a4-—b4 g3-h2:1 21 c2-c3 as—bs
9 d4-e4:1 d5-c5 h2-e1: 1 22 €3-d2:1 b§—cs5
10 c3-c4:1 bs-cs gi-f2:1 23 br-bz c§—C4
Ii d2-ca:i ¢5—bs (SB Oi 24 b2-a3
12 b3-b4:1 hs-is 70. f4—ts 22 e4-£4 2 1 and wins
13 e4-d4:1 e5-f4 20 f5-g5:1 =f4-€3. 1

7. TERRITORIAL CONTESTS

4.7.1. China: Wei-k’i; Korea: Patok; Japan: I-go, ‘the enclosing game’.
The favourite game of the upper classes in China, Korea, and Japan, and one
of the major board-games. Its age is often exaggerated; contemporary
references to it only become frequent under the Sung dynasty in China
(A.D. 960-1279), and it is significant that Chao Wu King, who lived between
90 W AR-GAMES [CH. 4
970 and 1127, records how he enlarged the existing Chinese chessboard by
dividing it lengthwise and across to produce the board of 19x 19 points on
which wei-k’i is now played (HC. 124). The game spread to Korea and thence
to Japan, where the first of the master-players of i-go whose name has been
recorded flourished 1465-1500. TheJapanese have far outstripped the Chinese
in their mastery of the game.!
Wei-k’i is played by two persons on the 361 points of an unchequered
lattice of 18 x 19 cells. The central point and the four corners of the central
square of 12X12 cells are marked in China, Korea, and Japan; in Japan the
middle points of the sides of this square, and in Korea the twelve points of
quadrisection of the sides of this square are also marked. None of these points
has any importance in the game, except that, when a player gives odds of a
number of men, these are placed on the marked points. Each player has 181
men, usually button-shaped stone counters, white for one player and black
for the other. The weaker player takes the black men and moves first, except
when white gives odds, when he begins. The players move alternately, enter-
ing one man at a time on any empty point (except in the case of a ko, men-
tioned below). After entry, no man can be moved again.
Each player’s aim is to secure territory, so that at the end of the game he
may have enclosed as much vacant space as possible. One or more men which

FIG 37. WEI-KI. VARIOUS CAPTURES

are compactly enclosed by enemy men are taken and removed at once from
the board. So long as an adjacent point in an orthogonal direction is unoccu-
pied, a man and any group of men of which it forms a part are free. Examples
of the minimum number of men necessary to enclose and take a man or
The game has a considerable literature in Japan. The best European works are O. Korschelt,
‘Das “Go” Spiel’ in MDGNVO., Aug. 1880, July 1881; S. Culin, Korean Games, 1895; Z. Vol-
picelli, “Wei-chi’ in JRAS., Chinese Branch, xxxi (1894); A. Smith, The Game of Go, 1908;
H. F. Cheshire, Goh, Hastings 1911; Ct. D. Pecorini, The Game of Wei-chi, 1929.
4.7.1] WAR-GAMES gI

group of men are shown in Fig. 37. A group of men which, while not
actually enclosed, will inevitably be eventually enclosed is regarded as dead,
and is left alone until the end of the game. Players soon learn to judge whether
a group is defensible or not and. will abandon the contest for it as already
settled and turn to other parts of the board. The ruling principle here is that
a group containing two disconnected empty points can never be taken (see

FIG. 38. THE WHITE GROUP IS SAFE FROM ATTACK

Fig. 38). Another situation which is left alone is that known inJapan as seki,
‘impasse’. It arises when a vacant space is surrounded partly by white and
partly by black in such a way that, if either player enters a man therein, his
opponent can thereupon capture the whole group. Both players naturally
leave such a position alone.
The game ends when the two territories are in absolute contact, or when
both players agree that no more territory can
be gained. The dead men are then removed a 6 c de f
by the player who would have won them if ied
the attack had been pressed, and these and 18
other men, that have been taken and re-
moved from the board, are used to fill up
points in the opponent’s territory. The player jg
who has enclosed the larger territory wins.
There is one exception to the procedure
when a single man can be taken which is 4
known in Japan as ko (see Fig. 39). In this
position, if it is white’s turn to play, he
enters 2 man on D16 and takes and removes FIG. 39. WEI-K’'l. KO
the black man on C16. Black is not allowed
now to reply E16, taking D16 (which would lead to a perpetual repetition
of moves), but must play elsewhere and can only play E16 after white has
92 WAR-GAMES [CH. 4
moved again. This is really the only rule in wei-k’i. Like draughts, the game
seems at first a very simple one, but this is an illusion about both games. It
will soon be found that wei-k’i demands far-reaching calculations, which
make it one of the most difficult games that man has invented.
4.7.2. Europe: Reversi. Played by two persons on the chessboard with
flat disks with one face red and the other black. When a man is captured, it
is not removed from the board, but is turned over to show that it now belongs
to the other player. Men are entered one at a time in alternate turns of play,
and once entered, cannot be moved. The aim of the game is to occupy the
larger number of cells. Men are captured by line interception. If it is Red’s
turn to play and a sequence of cells in row, column, or diagonal is occupied
from one end by first a red man and then a number of black and the next
cellis empty, Red, by entering a man on this empty point, takes the sequence
of black men and by turning them over adds them to his men. If the entry
of a man completes line interception in more than one direction, capture is
made simultaneously in all these directions.
Reversi was invented between 1880 and 1890.
8. BLOCKADE GAMES
4.8.1. India, Punjab: Do-guti, ‘two game’ (H. C. D. Gupta, d, 143).
Played on a diagram which consists of a square with its two diagonals, one
of the sides of the square being omitted. Two persons play, each having two
men, usually pebbles, which are placed as in Fig. 40. There are only five

Korea Punjab, China Siam

FIG. 40. KO-NO
points (the four corners and the intersection of the diagonals) on which the
game is played. A man can move along a line of the board to an adjacent
point which must be empty. No captures can be made and the aim of the
game is to blockade the opponent’s men.
4.8.2. China, Kwang-tung Province: Pong hau k’i; Korea: On-moul-
ko-no; Siam: Sua tok tong (Culin, a, 101). The same game as 4.8.1 with
slight variations in the form of the board as shown in Fig. 40.
* Culin, op. cit., gives three other ko-no games which are played in Korea, nei-pat-ko-no,
played on the points of a 4x 4 board; o-pat-ko-no, played on the points of a 5 x § board; and
ryouk-pat-ko-no, played on the points of a 66 board, but the descriptions of these are
unintelligible.
4.8.3] WAR-GAMES 93

4.8.3. Maoris, New Zealand: Mu-torere (Elsdon Best, 111-13; also in

Man, 1917, p. 7). Two persons play. The board (Fig. 41) is an eight-rayed
star with cells (kewai) at the ends of the rays and one at the centre (putahi).

FIG. 41. MU-TORERE (Maoris, New Zealand)

Each player has four men, placed, one side on 1-4, the other on 5-8. A man
can move either to the central point or to the end of an adjacent ray, pro-
vided the point is empty. No captures may be made, and the first player to
blockade his opponent wins. The first move must allow the second player
to play. Best gives the following example of play:
I. $9, 4-5; 2. 9-4, 3-95 3. 4-3, 9-45 4. 3-9, 2-35 5. 9-2, 4-9; winning.
Q. CLEARANCE GAMES
4.9.4. Europe: Solitaire. Played by one person on a board which in
Britain has thirty-two holes forming the cross on which the older game of
fox and geese is played, or in France thirty-seven holes (see Fig. 48). The
French board is also used in Sweden. One hole is empty and all the other
holes contain one marble. No marble can move except to make a capture
by the short leap of the draughtsman. The aim is to take all the marbles by
a series of captures except one, and this must be left in 2 named hole. The
game reached England in the eighteenth century.
4.9.2. England: Leap-frog. Any number of persons play on a board
which is divided into cells by parallels to the sides. The number of cells is
usually from 15X15 to 18X18. Every cell is occupied by a man. The
game is begun by each player in turn removing one man from the board;
after this has been done, every move must be a capture by the short leap of
the draughtsman, and if multiple captures are possible, these are made in
the same turn of play. The game ends when no more captures can be made,
and the player who has taken most men wins.
04 WAR-GAMES [GHs4

A more interesting game of this type was invented by me in 1898. Instead

of filling the cells of the board with undifferentiated men, I used counters
of four colours in the proportion of one green to two red, three yellow and
four white, and assigning different values to the colours, white counting 1,
yellow 2, red 3, and green 4. The game is played in the same way as with
undifferentiated men, but the player wins whose captures give the highest
total value.
4.9.3. Malaya and Indonesia: Chuki (Culin, d, 871-3, based on informa-
tion obtained from Hon. Dato Meldrum of Johore). Two persons play on a
board of 12X12 cells, the central point being occupied by a small raised
square or circle on which a bowl stands, and the three dice with which the
game is played are thrown in this bowl. Each player has sixty men, arranged
on the points of the board as shown in Fig. 42. Each player removes his men
in accordance with his throws.

FIG. 42. CHUKI (after Culin)

The game is mentioned in the Sejarah Malaya (early seventeenth century)

and Culin quotes a number of definitions from Malay dictionaries. It is
generally regarded as borrowed from China, and it may be identical with
a game fjeki, played in Amoy, which J. Pijnappel (Maleisch-Hollandsch
Woordenboek, 2e druk, 1875, i. 112) describes as a kind of Chinese hazard
game.

IO. WAR-GAMES PLAYED WITH LOTS OR DICE

Attempts have been made from time to time to use dice or lots to deter-
mine the moves in games of skill, but few of these have had more than a
temporary vogue. I have already mentioned the use of dice in the larger
4.9.3] WAR-GAMES 95

merels. More persistent were the attempts to play chess and its derived
games with the help of dice, although it was obvious that they destroyed the
essential quality of these games. As the Vetula (1. XXxiii) says:
Cum deciis autem qui primus lusit in illo
Foedavit ludum, languebit namque satelles
Immotus, nisi sors deciorum moverit ipsum.
Nec fuit hoc factum; nisi vel quia non nisi pauci
Ludere noverunt tractim; vel amore lucrandi.

The Indian four-handed chess was played with a long die from the eleventh
to the fifteenth century, and Muslims played oblong chess (4 x 16 cells) with
a cubic die from the ninth to the fourteenth century. The Alfonso MS. uses
a seven-sided die in its great chess (12 X 12 cells), and says that ordinary chess
may be played with a cubic die. A few medieval references show that chess
was occasionally played with dice outside Spain.
The only war-game played with lots that shows any vitality is tab which
is played by Muslims from Persia to West Africa and as far south as Somali-
land.
4.10.1. Egypt: Tab; Arabs: Tab al-qasab, tab wa dikk; Persia: Bazi
7548
qamish, ‘stave game’; Turkey: Qamish uyuni; Hanzoan: Sitta (Hyde, 217- YX )3
23; Lane, 317-20). Two persons play on an oblong board of four rows, each
containing the same odd number of cells (13 to 29 according to Hyde, 7 to 15
rckeyg2 Y
according to Lane). Each player has sufficient men to put one in each cell of his
outer row. Each player moves his men from left to right along his first and
third rows, and from right to left along his second and fourth rows (except as
described below). The moves are given by the throw of four staves. which
have one face plane (‘white’) and the other curved (‘black’). They are
thrown against a wall or a piece of wood stuck at a slant in the board or
ground, The throw is given by the number of whites falling upwards: none
up = 6; one up = 1 and is known as tab; two up = 2; three up = 3; and
four up = 4.! Each player goes on throwing until he throws 2 or 3, when
his turn of play ends. No man (kelb, ‘dog’, pl. kilab) can be moved until a
tab is thrown, which converts a ‘christian’ to a ‘muslim’. The first tab moves the
front man one cell forwards on to the second row. Following tabs must be
used in advancing the remaining men on the back row one step forward
until all have been moved once. When all the men have been moved, tabs
may be used on any row, or may be kept in mind until they can be used
more effectively. After reaching the second row, men may be piled and
then the whole pile, called “eggeh’ (Hyde, malik, “king’) is moved as if it were
a single man: a pile can be unpiled by a throw of tab, or can be taken as if it
were a single man. When a man reaches the second or third row, and reaches
the end of the row, it may retrace that row in the reverse direction if the
player likes to do so, but a pile must be first unpiled. When a man reaches
’ Hyde interchanges 4 and 6.
96 WAR-GAMES [CH. 4
the fourth row, it cannot be played so long as the player has other men
which he can play.
4.10.2. Somaliland, Darod tribe: Deleb (Marin, a, 508). Played by chil-
dren who are arranged to form two sides; all the children on one side play in
succession, and then all the children on the other side play in succession. The
board consists of three rows of any number of cells, and the men are placed
and play on the points of the board. Except for the opening play the game is
played in the same way as 4.10.1. The throws of the staves have the same
values as in 4.10.1, but the throw of tab is called deleb. The opening play is in
three stages before the first man is moved: (1) the player must throw deleb.
As soon as this is done, he goes on at once to the second stage. (2) He plays
with three staves; if he throws three black or two black and one white, he
adds a deleb to his score and continues to throw. If he throws three white, he
scores 4 and continues throwing. As soon as he throws one black and two
white, he passes on to the third stage. (3) He plays with two staves. If he
throws two black, his turn is ended and he loses all scores in hand of delebs
and 4’s, and has to start the first stage again. If he throws two white, he
scores 4, but has to return to the second stage. If he throws one black and one
white, he is free to begin to move. He moves his front man forward, using
all the delebs and 4’s in hand, and his first move ends.
4.10.3. Sudan: Sija(t) el taba (Davies, SNR., viii. 146). Played on boards
of 6X10 or 4X10 cells. Two players. Each player has six men which are
placed on the three cells at each end of his first row. The direction of play is
the same as in 4.10.1, from left to right on his odd-numbered rows and from
right to left on his even-numbered rows. The moves are given by the throws
of three staves: two black and one white = 1 (taba); one black and two white
= 2 (ydmin); three white = 4; three black — 6. Each player continues to
throw until he throws ydmin which ends his turn of play. A throw of taba is
necessary to move a man on the back row. When all the men have made
one move, throws of tab are kept in mind for future use. Men are taken as in
4.10.1.
4.10.4. Arabs of Palestine: Kidz, ‘walnut’ (Hyde, 223-4). Played on a
board of 4 22 holes, each player having eighteen men and four kings, but
without any difference of move or power, and the Jerusalem Arabs called
all the men indiscriminately ‘dogs’ (kelb, pl. kilab). The men are pegged for
insertion in the holes. Moves are given by the throws of four staves with the
same values as in 4.10.1, but the throw of 1 is called kidz. Played in the same
way as 4.10.1, except that men cannot be piled.
4.10.5. Moors of Senegal: Siga (Parker, 608, quoting Caillie, Travels
through Central Africa, 1830, i. 127). Played by two, four, or six persons who
constitute two sides. The board has 3 x 24 holes, and moves are given by the
throws of six staves (flat oval pieces of wood, black on one side and white
on the other). When all, or all but one, have the same face up, the throw is
called siga, scoring 1 and giving another throw. All other throws are valueless
4.10.5] WAR-GAMES 97

and end the turn of play. Men are represented by straws of two distinct
colours. Apparently they move one cell at a time when siga is thrown, and
enemy men are captured by playing a man to the square which they occupy.
The side wins which takes all the opponent’s men.
4.10.6. Dahomey: Awa(ng)du (K. C. Murray, who watched fishermen
from near Porto Novo, Dahomey, who were playing at Lagos in April 1951).
Played on the points of a board of 4x 11 cells, giving five rows of twelve
points on which the men are placed and moved, the throws of six cowries
giving the moves. Four people play, two throwing the cowries and two
moving the men. Each side has twelve men which are arranged one on each
point of the side’s back row. Moves are given by the throws of the cowries:
six backs up is called opoto; six backs down is ogo; all but one up or down is
siki, counting 1; two backs up and four down is ano, counting 4; three up
and three down is akrosan, counting 3; four up and two down is aviatu,
counting 2. The values of ogo and opoto were not noted, but apparently were
between 10 and 20 and could be divided to effect a number of captures in a
turn of play. A throw of siki, ogo, or opoto is necessary to begin the game.
One side moves left to right along its back and odd rows, the other side right
to left along its back and odd rows, so that both sides move in the same
direction along the inner rows. If a throw brings a man to a point occupied
by an opponent, the latter is taken. The game is won when one side is
reduced to one or two men.

Il. WAR-GAMES OF WHICH WE HAVE NO CERTAIN KNOWLEDGE

4.11.1. Chile, Aracaunian tribe: Comican (Molina, Civil Hist. Chile,
um. x; see Forbes, Hist. Chess, 1860, App. lvi). Said to resemble chess.
4.11.2. Java: Malingan, malen’gan, ‘theft game’ (Sir T. S. Raffles,
i. 350; D. M. Campbell, ii. 1017, which adds nothing to what Raffles says;
Plitschke, 193). ‘Played with eighteen pieces on squares. The object is to
surround your adversary’s pieces.’ Possibly a hunt-game.
4.11.3. Tibet: Mig-mang, ‘many eyes, i.e. cells’ (Sarat Chandra Das, 338;
cf. HC. 43 n., 367). Said to resemble chess, but more likely draughts.
4.11.4. Hawaii: Ko-na-ne (Culin, e, 243, who quotes various mentions,
none of which gives any clear indication of how it is played). There seems
no fixed size of board; Culin mentions boards of 10x15 and 14X17 cells.
Nor is there any fixed number of men, for which pebbles of two colours are
used. Capt. Cook (Voyage to the Pacific in 1776-80, ed. Capt. James King,
iii. 144) says that it is very much like our game of draughts, but more intri-
cate, and that the men move from square to square. Corney (Voyages in the
N. Pacific, 1813-18, Honolulu 1896, p. 106) calls it draughts, and Andrews
(Hawaiian Dict., 1865) defines it as a game like checkers, a species of pu-ni-
pe-ke (5.3.3.) and defines pa-hu-u-hi-u as the name of a game like ko-na-ne.

5356 H
 5
HUNT-GAMES
Most Asiatic peoples who play a war-game also play a hunt-game on the
same board. Hunt-games are played by two persons, one with many pieces
and the other with not more than four pieces, and the player with the larger
number of men endeavours to take his opponent’s men or to hem them in so
that they become immobile. The larger body represents a party of hunters
and the smaller body is their quarry, a small number of dangerous animals
which can kill a hunter who has lost touch with his comrades. In the older
games the quarry is a single animal and its capture is not difficult, provided
the hunters keep together and advance methodically. To remedy this defect,
the later Asiatic games have increased the quarry to as many as four animals,
all of which have to be cornered.
The hunt-game is a later development than the war-game and the earliest
example of a hunt-game occurs in the Alfonso MS. of 1283, where it is
included among the alquerque games and is played on the alquerque board
(Fig. 27). This suggests that the Spaniards obtained the game from the Moors
and that the game was an Arab game, although I have found no mention of
a hunt-game in the older Arabic literature. The diffusion of hunt-games in
Asia, and especially its spread to southern India and Indonesia, coupled with
their non-existence in Africa, supports an Asiatic invention.
Hunt-games achieved a certain popularity in France and England during
the Middle Ages and had their own special board, one obtained by joining
together five smaller merels boards to form a cross (see Fig. 47). Indeed, as
late as 1472-83, the game was still known in England by the name of merels,
though it was played with a fox and a pack of hounds. All the older hunt-
games were played on the points of lined boards, but the same tendency that
we have seen in the case of alquerque also affected the hunt-games, and the
powers of move of the hunters were reduced to the orthogonal directions
or to the forward directions only. In Europe from the sixteenth century
onwards, hunt-games were transferred from lined boards to the chequered
chess-board, on which they used the cells of one colour only as in draughts.
The same change has happened in some parts of Asia.
I classify hunt-games by the type of board on which they are played:
(1) games played on the alquerque board; (2) games played on enlarged
alquerque boards; these classes contain some Chinese and Japanese games
which are war-games with unequal sides; (3) the older forms of fox and
geese; (4) the modern games of fox and geese and asalto; (5) leopard games
played on triangular boards; (6) tiger games, which I arrange by the number
of tigers.
 5.1.1] HUNT-GAMES 99

I GAMES ON THE ALQUERQUE BOARD

5.1.1. Spain: De cercar la liebre, ‘catch the hare’ (Alf. 916). Played on
board A (Fig. 43), the men arranged as in the diagram, the hare on the central
5

Ww

fo

CG
A B GC
FIG. 43. THE HARE GAME

point, and the twelve hunters on the first and second rows and on the end
points of the third row.
Both hare and men have the same power of move, one step along a marked
line in any direction, the point reached being empty. The power of move
accordingly varies from point to point. The hare alone can take, and captures
are made by the short multiple leap of the English draughts. The aim of the
game is to enclose the hare so that it can neither move nor take. Alf. says that
this is possible even if the number of hunters is reduced to ten, and that a
good player can give his adversary the odds of one or two men.
In course of time the board was simplified by the omission of some or all
of the slant lines with a corresponding reduction on the power of move.
Culin (g, 798) says that the juego de la liebre is now played in Spain on board
C. Intermediate forms are shown in the following American Indian games.
5.1.2. Piman stock, Papago Indians, Arizona: Pon chochotl, ‘coyote and
chickens’ (Culin, g, 794). Both Papago and Mexicans play on board A with
the same arrangement of men, using a red bean for the coyote and grains of
corn for the chickens.
5.1.3. Tanoan stock, Tewa, New Mexico: name unrecorded (Culin, g,
797). Played on board A with the same arrangement of men.
5.1.4. Mexico: Coyote (Culin, d, 876). Played on board B with twelve
men.
5.1.5. Tanoan stock, Tigua Indians, New Mexico: Ko-app-paw-na,
‘Indian and jack-rabbits’ (Culin, g, 798). Played on board C (whee-e-na).
Small stones are used for the rabbits (kd-na), but the opponent merely points
with a stick to the position occupied by the Indian (tu-na-mah).
Board A (Fig. 43) is also used for some Asiatic hunt-games.
5.1.6. Japan: Yasasukari musashi, ‘the soldiers’ eight-way hunt’ (Culin,
a, 77; cf. Himly, a, 481). Played on board A. One player has a rikishi,
‘general’, placed on the central point; the other has sixteen musashi, ‘soldiers’,
100 HUNT-GAMES [CH. 5
placed on the sixteen points round the edge of the board. Culin obtained the
figure of the board from Simayosi Anko’s Sam sai dzu e, a translation and
revision of the Chinese encyclopedia, San-t’sai-y’u-hwei, which Simayosi
Anko completed in 1712, but says nothing on the method of play. The game
now seems to have been replaced by juroku musashi (5.2.3).

2. GAMES ON ENLARGED ALQUERQUE BOARDS

5.2.1. Peru: Solitario (Culin, d, 876, who says it is a game resembling the
Malay tiger games). Played on the board of Fig. 44.
5.2.2. China: Shap luk kon tseung kwan (Culin,
d, 874); shi liu kan tsiang kiin, ‘sixteen pursue the
general’ (Himly, a, 469). Played on board A, Fig. 45.
One player has sixteen soldiers and the other a general,
placed as in the figure. All the pieces have the same
move, one step along any marked line of the board,
but the general alone can enter the triangle (Culin,
mau ts’z, Himly, maosse, ‘privy’), and if he is confined
to the triangle, he loses. Both general and soldiers can
FIG. 44. SOLITARIO. capture, but in different ways. The general takes by
PERU intervention (see p. 11) by occupying the point im-
mediately between two soldiers on the same marked
line, when he takes both; the soldiers by interception (see p. 10) by occupy-

A B
FIG. 45. SIXTEEN SOLDIERS (China andJapan)

ing the points immediately on both sides of the general, both soldiers and
general being on the same marked line. Thus, in the position diagrammed,
 5.2.2] HUNT-GAMES IOI
if the soldiers play first and move e2-d2, the general may reply c3-b2,
taking the soldiers on a3 and cz by intervention; but now the soldiers may
move d2-c2 and take the general by interception. Since the interception
capture is more primitive than the leap capture in war-games, this may be
an older type of hunt-game than the others which I describe.
Himly says that the board is often marked out on roads in China, and the
game is often played by labourers and children.
5.2.3. Japan: Juroku musashi, ‘sixteen soldiers’
(Culin, d, 873; Prof. Tsuboi). Played on board B,
Fig. 45, which is copied from a board (no. 7090)
in the Museum of Archaeology, University of
Pennsylvania. When Prof. Tsuboi drew the board
for me, he omitted the two shorter lines in the
triangle. One player has a general (taisho) and
the other sixteen soldiers (musashi) which are
arranged as in the diagram. All the pieces have
the same move, one step along a marked line
through the point on which the piece stands, but
the general alone can enter the triangle, and if he
is confined to it, he loses. The general alone can
capture, and he does so by the short leap of the
draughtsman. The general wins if he takes all
the soldiers, and the soldiers win if they reduce
the general to immobility or confine him to the
triangle.
5.2.4. China: Yeung luk sz’ kon tseung
kwan, ‘General against twenty-six rebels’ (Hyde,
FIG. 46. YEUNG LUK Sz 214, who obtained the game from Shin Fo of
KON TSEUNG KWAN Nanking). The board and arrangement of the
pieces are shown in Fig. 46. The general (tseung)
occupies the central point of the rectangle and the twenty-six rebels (kwan),
the forces of a freebooter named Yeung luk sz’, occupy the remaining points
on the three middle columns. All pieces move one step at a time in the
orthogonal directions. Captures are made as in 5.2.2, the general by inter-
vention and the rebels by interception. The game is probably now obsolete.
3. Fox AND GEESE. THE OLDER GAME
5.3.1. Britain: Fox and geese; dialectally, Tods and lambs (Gomme,
ii. 298); France: Renard et les poules, also Marelle quintuple (Moulidars) ;
Italy: Lupo e pecore; Germany: Fuchs und Ganse or Hiihner, Fuchs im
Hiihnerhof; Holland: Schaap en wolf; Sweden: Rafspel; Iceland: Ref-
skak; Russia: Volki ovtsy (Fiske, 146-56).
An older Icelandic name would seem to have been hala-tafl, ‘fox game’
which is mentioned in the Grettis saga (Fiske, 156), and Du Chaillu (ii. 354)
102 HUNT-GAMES [CH. 5
gives a drawing of a game-piece in the form of an animal which was found
in Norway.!
References to fox and geese in literature are scanty, and I have only found
one before the seventeenth century. This occurs in the accounss of the Royal
Household, temp. Edward IV, when ‘two foxis and 26 hounds of silver
overgilt’ were purchased to form two sets of “Marelles’ (V. B. Redstone,

FIG. 47. FOX AND GEESE

‘England among the Wars of the Roses’ in Trans. RHS., N.S., Xxi, 1902,
pp- 195 ff., where ‘26’ is printed ‘46’ in error). It is possible, therefore, that
some of the many references to merels in medieval literature may refer to
this game. Two mentions in works of the seventeenth century are in Shackley
Marmion, A fine Companion, 1633, 0. v: ‘Let him sit in the ship . .,. and play
fox-and-geese with the foreman’ and in Lovelace’s verses prefixed to Beale,
Giochimo, 1656: “Men that could only fool at fox-and-geese are new-made
polititians by thy book.’
That fox and geese was played before the fifteenth century is evidenced by
the occurrence of boards incised on the stone seats in the cloisters at Glouces-
ter Cathedral. Here the component squares only contain the diagonals,
though one diagram adds the missing lines to the central square. A complete
board, scratched on a flat stone, was retrieved from the well in Norwich
Castle, and the board of Fig. 47 also occurs in the cloisters of San Paolo
without the walls, Rome (Miss K. M. E. Murray).
In its earliest form the game was played with thirteen geese arranged as in
Fig. 47, and a fox which was placed on any of the vacant points (but usually
on the central point, this being considered its strongest position). All the
pieces had the same move, one step along any marked line through the point
occupied. The fox alone can take, which it does by the short leap over a
goose on an adjacent point in the direction of move to the point immediately
t John of Salisbury (Politicraticus, c. 1156) includes volpes in a list of dice-games which also
includes some technical terms for the throws of the dice, but this probably precludes volpes from
being the name of a hunt-game.
5-3-1] HUNT-GAMES 103

beyond the goose, and a series of similar captures may be made in a turn of
play. The geese win if they can hem in the fox so that it cannot move; the
fox wins if he takes so many geese that too few are left to enclose him. With
reasonable care the geese should always win.
The original form of the game is still played in rural Britain, e.g. in
Lincolnshire and Shropshire (The Times, 23 Nov. and 6 Dec. 1939) and in
Iceland (Isl. Gatur, 298) where it was very popular in the eighteenth century.
After 1600, changes in the game appear. These were adopted locally with

OUZOL=O

x0 OLnx
OF OLD OO se 3O mr)

Oe Oe) awn Gee)

© art, Oe O) Rak OPS il)

Xe O22 OOF. &

Or
Oe to
FIG. 48. SOLITAIRE BOARD

the result that there is now no uniform code of rules in any country. The
geese were deprived of their power of moving diagonally or backwards. It
would seem that this made the geese too weak, for their number was in-
creased, first to fifteen by adding two geese on A, A, and later to seventeen
by adding two more geese in England on B, B, or in France on C, C, in Fig.
47. In some games the fox’s power of move was also reduced to the orthogo-
nal directions but his power of taking diagonally continued. The abandon-
ment of diagonal moves led, as in alquerque, to the removal of the slant
lines from the board, and, after the invention of solitaire, to the transfer of
the game to the solitaire board of thirty-three holes in England or of thirty-
seven in France (see Fig. 48). I add a brief summary of the modern rules as
given by Fiske and in such game manuals as I have consulted.
1668. France (Maison des Jeux académiques). Thirteen geese which can
move neither diagonally nor backwards. The geese play first.
1681-2. England (Randle Holme, a, 67). Also in C. H. Bennett, Games of
Skill and Conjuring, London & New York 1861. Fifteen geese.
1830. England (Strutt, rv. ii). Seventeen geese moving orthogonally but
not backwards. “The great deficiency of this game is, that the fox must
inevitably be blocked if the game is played by a skilful hand; for which
reason, I am told, of late some players have added another fox.’ This is
probably asalto (5.3.7).
104 HUNT-GAMES [CH. 5
1839. France (Moulidars, who also gives the original game). Seventeen
geese, moving orthogonally but not backwards. Some players only allow
the fox to move orthogonally, but this gives him no chance; others give him
his original move but limit his power of capture to the orthogonal directions.
Some players allow the fox to take several geese in succession, but it is
better to restrict him to a single capture on a move. If the fox omits to take
when he can, he is huffed, i.e. his opponent enters a new goose on his back
row as soon as a point on this row is available.
1839. Sweden (Handbibliothek for Sallskaps-ndjen). Played on the French
solitaire board with twenty-two geese on the four lowest rows. Both fox and
geese move orthogonally, but the geese cannot move backwards. Fiske thinks
this is the French game and that the original game is more frequently played
in Sweden.
1868. England (Boy’s Own Book, 627). Seventeen geese moving orthogo-
nally but not backwards; the fox also moves orthogonally only.
1935. England (Encycl. Sports, Games and Pastimes). Seventeen geese. All
moves are orthogonal but the geese cannot move backwards. The fox, by
agreement, may also move diagonally. Inexperienced players often use the
original powers of move. If the fox omits to take when he can, he is huffed,
i.e. his opponent enters a goose on his back row if a point on it is empty; if
none is empty, the penalty lapses.
The original game with thirteen geese has spread to the following lands:
5.3.2. Canada; Cree and Chippeway Indians in Assiniboia: Musinaykah-
whanmetowaywin (Culin, g, 791). Played with a king (oke-mow, or musi-
nay-khan-whan) and thirteen men.
5.3.3. Hawaii: Ma-nu (Culin, e, 244). One player has thirteen bits of
stone and the other points with a stick to the point occupied by his piece
(pu-ni-pe-ki, which Culin suggests is a native rendering of Bonaparte, pointing
to an introduction in the early nineteenth century).
Enlarged games have been played on larger cruciform boards with the
same system of lines connecting the points and defining the powers of move.
5-3-4. England: Double fox and geese (R. Holme, a, 168). “There is
another sort of board called the double fox and geese, it has twice as many
holes as this (Fig. 47), in which is played two foxes and thirty geese.’
5.3.5. England: Treble fox and geese (R. Holme, a, 168). ‘In which
there is three or four foxes and fifty or sixty geese. Every square in the
double board hath in this another crosse line made in it from corner to corner
of the square.’
5.3-6. Italy: Lupo e pecore. A sixteenth-century compendium of games
from Venice in the Victoria and Albert Museum (no. 154, 1900) includes the
board shown in Fig. 49.
5.3-7- Spain, England: Asalto; France: Assaut; England: Officers and
men; Germany: Festungsspiel, Belagerungs-spiel; Denmark: Belejring-
spel; Sweden: Belagringsspel, Fastningsspel; Iccland: Beleiringsspil
5-3-7] HUNT-GAMES 105

(Fiske, 151-3; Brunet y Bellet, 190, 205; Tressan, 153; Tom Wilson, 201-5).
Many of these names are drawn from manuals which have derived most of
their contents from French manuals. I doubt whether it has ever been widely
played in any of the countries enumerated.
Played on the board of Fig. 47, the upper arm of which is surrounded by
a double line to represent a fortress within which one player puts his two

INA
VAINVANVAINGA
Sa

FIG. 49. FOX AND GEESE. VENICE

officers. His opponent has twenty-four soldiers who occupy all the points
without the fortress. All pieces move one step along any marked line but the
soldiers must always move towards the fortress. The officers alone can take
by the short leap, and if an officer fails to take when he can, he is huffed, i.e.
removed from the board. The soldiers win when they occupy every point
within the fortress or when the officers are hemmed in and are unable to
move.
4 Fox AND GEESE ON THE DRAUGHTSBOARD
Following the rise in popularity of draughts in the sixteenth century,
players in Italy began to play hunt-games on the draughtsboard. A Latin
manuscript of this century in the Communal Library, Perugia (L. 27, ff. 1636,
164b, 165a), gives three diagrams (without explanatory text) of games of
this kind.
5.4.1. Italy: Ludus rebellionis (Perugia MS.). The first diagram shows
the following arrangement of men: Kh8, Qb8, sixteen Pawns on the alternate
cells (starting from ar) of the first four rows. There are marks on £8, h8, and
h6, of which the purpose is unknown. The second diagram has the same
arrangement in black, and in red, Khr, Qb1, and sixteen Pawns on the alter-
nate cells (starting from bs) of the fifth to eighth rows, and marks on f1, hr,
and h3. The third diagram has the same arrangement of the black men but
106 HUNT-GAMES [CH. 5
the red are now Ka8, Qg8, sixteen Pawns on the alternate cells (starting from
br) of the first to fourth rows and marks on a8, c8, and a6. These diagrams
seem to be varieties of fox and geese with two and four foxes.
5.4.2. France: Jeu de Renard (Mallet, 445-7). Played on the white cells
of the chessboard with one fox and twelve hens on the white cells of the first
three rows. All pieces move one step diagonally, the fox forwards and back-
wards, the hens forwards only. The fox alone captures by the short leap.
The hens try to enclose the fox so that he cannot move. This they can easily
do, even if reduced to nine only. Mallet goes on to say that some players
play with two foxes who must move alternately, and from thirteen to
sixteen hens, placed on the lower four rows. This appears to be the first of
the Perugia games, 5.4.1.
5.4.3. India: Shatranj shir bakri, ‘lion and goats chess’ (Lala Raja Babu,
195). Played on the chessboard between two lions and thirty-two goats, the
latter arranged on the cells of the first four rows.
5.4.4. England: Fox and geese; Sweden: Vargen och faren; Ceylon:
Koti keliya (Parker, 585). Played in England on the chessboard, in France
on the Polish draughtsboard of 10x Io cells, and in Ceylon on the 12x 12
board, using cells of one colour on chequered boards and on alternate cells in
Ceylon. Played with one fox (in Ceylon, a leopard), and one row of geese
(in Ceylon, dogs or cattle) who are placed on the first row. All pieces move
one step diagonally, the geese forwards only (in Ceylon, the leopard can
move two steps at a time). There are no captures. The fox wins if it pene-
trates the line of geese, the geese if they hem the fox in so that it cannot move.
Apparently a modern game in Europe, as it is not mentioned in Strutt.

5. THE LEOPARD GAMES

I use the name ‘leopard game’ for a group of games which are played on
boards of triangular outline in India, Siam,
Ceylon, and Indonesia. The simplest form of
board is that shown in Fig. so.
5.5.1. Siam: Len choa (Capt. Low, 380).
Played on the board of Fig. 50, one player having
a tiger, and the other six counters, which are
placed on the points of the board as the player
chooses.
5.5.2. Ceylon, coastal regions only: Hat
diviyan keliya, ‘the seven leopards game’
(Ludovisi, 33; Parker, 580). Played on the ten
FIG. 50. HAT DIVIYAN
KELIYA
points of Fig. 50. One player has a tiger and the
other seven leopards which are commonly known
as ‘dogs’. The tiger is usually placed on the apex of the triangle and the
second player enters his leopards, one at a time, in alternate moves with
moves of the tiger. All the pieces move in the same way, one step along
5.5.2] HUNT-GAMES 107

a marked line, but the leopards cannot be moved until all are entered. The
tiger, which alone can capture, takes leopards by the short leap. The tiger
wins if he takes so many leopards that they cannot confine him, the leopards
if they succeed in reducing the tiger to immobility.
Enlarged boards are used for the following games, and the number of
tigers is increased to three or five, and the leopards to fifteen.

A B
FIG. §I. PULIJUDAM

5-5-3. India, all parts: Pulijudam (Hindustani), ‘the tiger game’ (Culin,
d, 876). Played on Board A (Fig. 51). The three tigers are placed on the
marked points. The fifteen leopards are known as ‘lambs’. Played in the
same way as 5.5.2.
5.5.4. India: Rafaya (Hindustani); Ceylon: Demala diviyan keliya,
‘the Tamil leopards game’ or Koti sellama (Parker, 581). Played on both
A and B (Fig. 51). Three tigers and fifteen ‘dogs’. Played in the same way as
Oey
: ae Achehn, Sumatra: Meurimiieng-rimiieng-do’ (Snouck-Hur-
gronje, ii. 204). Played on Board A (Fig. 51) with five tigers and fifteen
sheep. Played in the same way as 5.5.2.
6. THe TIGER GAMES
I include under this name a group of hunt-games which are played in
Assam and other parts of India, Burma, Siam, Malaya, and Indonesia on the
enlarged alquerque boards (Figs. 28 and 29).! These are the favourite games
of the Malays. The general opinion is that their immediate source is southern
India (Snouck-Hurgronje, ii. 203; R. J. Wilkinson, 57). Unfortunately, few
writers who mention these games have troubled to give an exact description
of them. The number of tigers varies from one to four, and the number of
goats from eleven or twelve to twenty-three or -four. The general aim is to

! There is some overlap between this section and the first two sections of this chapter which is
partially due to lack of exact information. In the first two sections I have restricted the games to
those played in Europe and China and to games that are derived from these. In the present section
I deal with similar games which are apparently unaffected by European or Chinese influences.
108 HUNT-GAMES [CH. 5
hem in the tigers so that they cannot move, but the arrangement of the
pieces, the method of play, and the rules of capture vary from game to game.
I classify the games by the number of tigers.
(2) ONETIGER. 5.6.1. Malaya: Rimau-rimau, ‘tigers’ (Plitschke, IAFE.,
iii. 1890, p. 189). Played on the board of Fig.
52, which shows the notation that I employ
for the score of the illustrative games.

NWA The board is generally cut out of black or

white cloth, and the lines are worked in red.

IVAN Sometimes the board is of wood, and if no

VAAN
board is available, the Malays will trace one on
the ground. The two triangles are called
gunung, ‘mountain’. There are no_ special
VAS pieces; bits of wood, pebbles, or coins are

VASA
used, and if coins, the winner retains those

WAS / iAi Nee 5

laid down by the loser.
One player has the rimau, which is placed
on the vertex of either triangle; the other

Views. player has twenty-four orang-orang (pl. of

kk orang, ‘man’), nine of which are placed on
the nine central points of the square. All the
FIG. §2. RIMAU-RIMAU pieces move one step along a marked line. The
rimau alone has the power to capture, but
capture is not compulsory. If along a marked line the rimau has next to it an
orang or a succession of an odd number of orangs, and the point immediately
beyond the orang or row of orangs is empty, the rimau leaps over the enemy
pieces to the vacant point beyond and takes them.
The game begins by the rimau removing from the board any three orangs,
whether they form a row or not, and placing himself on any vacant point.
His opponent for the next fifteen moves enters his remaining orangs as the
play suggests, the rimau moving and taking in his turns of play. During this
preliminary play, no orang may be moved, but when all are entered, they
acquire and exercise their power of move. The player of the orang generally
gives up when he is reduced to ten or eleven men.
Plitschke gives the following example of play: the rimau is placed on c3 and
nine orangs on d2, d3, d4, e2, e3, e4, f2, £3, £4. The rimau removes €2, €3, e4
and is placed on e3. The game proceeds from this position:
Re O. R. O. TR O. Te. O.
t (Re3) 23 qa Rdt' tr “13°Rdse “fs 19 Rb_~ es-d4
2 Re3 rds" eF 8 Rd2 es 114 Rd4 33 20 Rb3 a3-a
3 Reg:-d4 24 ‘9 Rez -e, 15 Req b 21 Rb aa-a3
4 Reg gs 10 Rd3 bb 16 Ra:b b3-c3 22 Rb3 a-b
5 Re gt “ai Rey “aap ary, Ro e4—d4 and wins
6 Rerzd2. ez’ 12 Rd4’ 63. “18.Rbz
5.6.2] HUNT-GAMES 109

(b) ONE OR TWO TIGERS. 5.6.2. Simalur, an island off the north-west
coast of Sumatra: Rimoe (E. Jacobson, 80). Played on the board of Fig. 52.
One or two tigers placed on e3; eight goats placed on da, d3, d4, e2, e4, £2,
£3, £4, and sixteen in hand to be entered one at a time in succession until all are
entered. Tigers can leap over an odd number of goats but not over an even
number, and when played with one tiger only, it can only take one goat at
a time.
(c) TWO TIGERS. 5.6.3. Menankaba, Malaya: Main tapal empat (O. T.
Dussek, 71). Played on board A (Fig. 43). Two tigers (harimau, ‘tiger’) are
placed on the central point of the board, and the opponent has twenty-four
kambing, ‘goats’, in hand, which he enters one at a time in alternation with
the moves of a tiger. A tiger can move any distance in a straight line as
marked on the board, all the points passed over and the point to be occupied
being empty. A goat moves one step only along any marked line. The tigers
take by the short leap.
The player with the goats begins, and usually enters a goat on c2 which
one of the tigers then takes by c3-c1. This is done to separate the tigers.
5.6.4. Malaya (played by the Peninsular, Deli- and Menang-kabau-
Malays): Rimau-rimau (Plitschke, 189-90). Played on board shown in
Fig. 52. One player has two tigers and the other has twenty-two men. The
moves are the same as in 5.6.1, but the tiger can only take one man at a time
by the short leap.
At the beginning of the game, eight men (orang-orang) are placed on the
eight points surrounding the central point of the square. The two tigers may
be placed on any two vacant points or both can be placed on the central
point, e3. The game is opened by one tiger taking any single orang from any
point he likes, and then placing himself on any vacant point. The game is
then played like 5.6.1. Plitschke gives the following example. The tigers are
placed on e3 and g3 and the men on dz, d3, d4, e2, f2, £3, £4. The player with
the tigers then removes the man on f3 and places the tiger from g3 on c3
and the player with the men begins.
‘pees e5 15 d3-d1:d2 e4-d4 29 gi-f1 c3-d4
2 c3-e1:d2 3 16 di-d2 c1-d1 30 f1-er d4-e5
3 e3-c5:d4 63 17 d2-c3 gI-g2 3I 24-25 bb-c3
4 c5—d4 23 18 el-ci:dr_ f2-er 32 e3-f1 aa—bb
Sidt-d2 dz. 9i3 19 c1-d2 g2—f2 33 f1-g1 c3-d2
6 d2-bb:c3 ds 20 C3-C4 e5—€4 34 gi-f1 bb-c3
7 bb-c3 aa 21 ¢4=C5 ds-e5 35 f1-e1 a3-b3
8 c3-d3 bb 32 .05—a5 e5—f5 36 er-d1 c3-d3
9 d3-c3 a3 23 ds-es d4-cs 37 g5-fs 83-24
10 c3-b3 gi 24 es-g5:f5 fi1-g1 38 di-er 22-93
11 b3-c3 b3 25 d2-di e1-f1 39 e1-c3:d2 —_b3-a3
12 c3—b fI 26 95-24 gI-g2 40 c3-d4 e2-d2
13 b-c3 d2 27 di-er c5—ds 41 f5-g5 aj—a
14 c3-d3 cI 28 el-gi:f1 b3-c3 42 g5-f5 84-85
110 HUNT-GAMES [CH. 5
43 d4-c3 f2-e1 §0 e1-dr e3-d3 $7 fa-dase3 d3-d2
Aa C3 €3-€2 $r di-—cr f2-3 58 d4-e5 92-93
45 C2-CI e4-€3 $2) CI-C2 c3-d4 59 ¢3-f2 23-24
46 ci-di e1-f2 $3 C2-¢3 d2-c1 60 f2-g3 e4-€3
47 di-e1 d3-<3 $4 c3-d2 ez-€1 61 93-f2 ©3-¢4
48 e1—-di e3-d3 $5 d2-e2 d4-e4 62 f2-€3 Resigns
49 di-e1 d3-€3 56 e2-f2 g3-g2

5.6.5. Achehn, Sumatra: Meurimuéng-rimuéng (Snouck-Hurgronje, ii.

203). Played on the board of (Fig. 52), but the rules are a combination of those
of 5.6.1 and 5.6.3. Thus, the two tigers are placed initially on ¢3 and eight
sheep are placed as in 5.6.3. The tigers take as in 5.6.1. The player with the
sheep enters one of his sheep in hand whenever the tigers make a capture.
5.6.6. Java: Machanan (Sir T. S. Raffles, i. 350; Veth, Java, 1875, i. 625;
D. M. Campbell, ii. 1017; Veth and Campbell merely repeat what Raffles
says). According to Raffles, played by two persons, each of whom has a
tiger (machan);there are also eighteen (Veth says twenty-three) cows on the
board, and the player who captures the majority of the cows wins. I suspect
that Raffles is wrong in giving each player a tiger and leaving the ownership
of the cows with neither player. Other writers on Malay games (e.g. Wilkin-
son, 57) equate machanan with the other tiger games.
5.6.7. Celebes: Mamatyang-matyang, ‘tigers’ (B. F. Matthes, Boegi-
neesch Woordenboek, i. 256, quoted in Kaudern, iv. 311-13). Played with two
tigers and twenty-four men. Kaudern says that this game will always be
found in the coastal villages where Orang Bugis have settled, but that it
appears to be unknown in the interior of the island.
5.6.8. Ceylon: Diviyan keliya, ‘leopards’ game’, or Diviyalliya,
‘leopards’ square’, or Kotiyo saha harak, ‘leopards and cattle’ (Parker,
581-2). Played on the enlarged alquerque board with four triangles, Fig. 29.
One player has two leopards and the other twenty-four cattle. All the pieces
have the same power of move, one step along a marked line. The cattle
cannot take; the leopards take by the short leap, but only one beast at a time.
The game begins by one player entering a leopard on any point of the
board, usually the central point. Then his opponent enters one of his cattle.
Next, the second leopard is entered. The leopards are now free to move, but
the opponent goes on entering his cattle, one on each move until all are
entered, and only then may a beast make a move. If the leopards succeed in
taking eight cattle, they are almost certain to win.
5.6.9. India, Lower Bengal: Bagh bandi (J. M. Datta, 167). Played on the
board of Fig. 43a. Two tigers, placed on a3, c3, and twenty goats, placed in
piles of five on ba, b4, da, d4.
5.6.10. India, United Provinces, Karwi Subdivision: Bagh guti (E. de M.
Humphries, 117). Played on the board of Fig. 43c. Two tigers and twenty
goats, placed as in 5.6.9. Tigers can only take one goat each move, no matter
how many goats are in the pile leaped over.
§.6.11] HUNT-GAMES III
5.6.11. India, Punjab: Sher-bakar (H. C. D. Gupta, d, 143). Played on
the board of Fig. 43a. Two tigers, placed on a3, e3, and nineteen goats,
placed on ba, b4, dz, d4, in piles of five on three of these points and four
on the remaining point.
5.6.12. Manipur, Assam: name unrecorded but meaning ‘tigers and men’
(T. C. Hodson, 62-63). Played on Fig. 52, but Hodson omits the continua-
tions of the central vertical line over the two triangles. Two tigers and twenty
men; the tigers are placed on e1 and es; the men are placed in piles of five on
d2, d4, f2, and f4. A tiger kills a man each time he jumps over a pile to the
point beyond (the short leap). Presumably the piles of men are gradually
unpiled by moving the men singly one step along a marked line.
5.6.13. Siberia, Upper Yenesei valley: Bouge-shodra, ‘boar-chess’
(E. V. Savenkov, 25-26). Played with two boars and twenty-four bouza,
‘calves’. Nothing is said about the board, but there is one in the Minusink
Museum.
5.6.14. Mongolia: name unrecorded (G. Sdéderbom, ‘A Mongolian Game
Norobo’, in Ethnos, xv, 1950, p. 95). ‘The wolf-and-sheep game is played
on a chequered board or on squares drawn on the ground.’
(d) THREE TIGERS. 5.6.15. India, Deccan and southern India: Bagh-
bukri, ‘tigers and sheep’ (Herklots, App. viii). Played by children on the
board of Fig. 52 with three tigers and thirteen sheep.
5.6.16. Burma: Legwet kya (Shway Yoe, 278). Played with three tigers
and eleven or twelve little tigers.
(ce) FOUR TIGERS. 5.6.17. Assam, Sema Nagas: name unrecorded (J. H.
Hutton, b, 110). Played on board A (Fig. 43). Four tigers, one on each cormner-
point of the board. The opponent has twenty goats in hand, which are
entered one at a time in alternation with the moves of a tiger. Each tiger
moves in succession in the same rotation until all the goats are entered; after
this, any tiger can be moved. The moves of the tiger and goat are as in §.6.1;
the tiger only can take by the short leap.
5.6.18. India, Deccan and southern India: Mudrunggum (Herklots,
App. vii). Played by the lower classes with four tigers and sixteen sheep.
5.6.19. Malaya: Main rimau or Main rimau kambing (W. W. Skeat,
487). Usually played with four tigers and a dozen goats.
5.6.20. Siam: Len cua kin ngoa (Capt. Low, 379, who gives the follow-
ing account):
The game of ‘the tigers eating cattle’. In this game there are four tigers and twelve
oxen. The board has sixteen squares (? the simple alquerque board, Fig. 43a). The
tigers are placed at the four corners, an ox is placed on one square, and the nearest tiger
moves first. The tigers take by leaping over the heads of the oxen to the open square
behind them and not diagonally.
The tigers are taken and put off the board when hemmed in so that they cannot
move, or they are taken then by the oxen moving (only then) diagonally. If only two
tigers are left on the board, the oxen are victors.
112 HUNT-GAMES [Chas
There is another game called the sleeping tigers, and ten oxen are placed on a
fifteen-checked board. The tigers wait until they can take by leaping over the heads
of the oxen one at a time to a blank square—or by getting between three or four,
they take.
(Like so many of Capt. Low’s descriptions of games, this is too vague to
make anything of it.)
5.6.21. India, Orissa: name unrecorded (H. C. D. Gupta, c, 211). Played
on the board of Fig. 434. Four tigers and twenty goats.
5.6.22. India, Assam and Sikkim: Bhagchal, Bhagchakar, or Chah-
rachel (S. L. Hora, 2). Played on the board of Fig. 43c. Four tigers, placed
on the corner-points of the board, and twenty goats which are entered one
at a time and cannot move until all are entered.
The tiger game is played in India on other boards.
5.6.23. India, Central Provinces: Kaooa (H. C. D. Gupta, c, 165). Played
on the pentagram (Fig. 14), with one tiger and seven kaooas which try to
hem in the tiger. The tiger takes by the short leap.
5.6.24. India, Sikkim and Assam: name unrecorded (S. L. Hora, 2).
Played on the same board as gol-skuish (4.2.41), seven concentric circles and
six diameters, with one tiger, but the number of goats is not specified.
 6
RACE-GAMES
Race-Gamgs have been played from the earliest times for which we possess
records of games. The general characteristics of these games have been
summarized in Chapter 1, and the oldest race-games are discussed in Chapter
2. Today they are widely played by Europeans and their descendants in
America, Africa, and Australia, by most Asiatic peoples, and by the native
tribes of America. In Europe, the U.S.A., and Japan the invention of new
race-games continues every year. The Encyclopédie méthodique (1792) already
describes seven such games! and Culin (d, 843-8) gives a list of some 140
modern race-games already in the Museum of Archaeology, University of
Pennsylvania, in 1898. In the present work these games will be ignored: they
are mostly short lived and variations of the game of goose (6.4.5).
The number of race-games to be included in this chapter is still so con-
siderable that some system of arrangement is desirable, and I adopt the
following:
(a) Games of the Old World: 1. Oriental games affiliated to the Roman
alea (2.8.3). 2. The European games of tables. 3. Games on square boards.
4. Games on cruciform boards. 5. Single-track games. 6. Promotion games.
(b) Games of the New World. 7. Games played by the American Indians.
(c) The Arabic astronomical game.

I. ORIENTAL GAMES AFFILIATED TO ALEA (2.8.3)

6.1.1. Ancient Rome: Alea or tabula; Byzantine Empire: Tablé. This
has been treated in Chapter 2.
6.1.2. Persia, Islam, Babylonian Jews: Nard, nardshir (Hyde, 1-64).
Played on the backgammon board? by two persons, each with fifteen men,
the men, the moves being given by the throw of two dice, not three as for
alea. The modern Persian game (Himly, b, 679) starts from a fixed arrange-
ment of the men. One player (C) has two men on a, five on m, three on r,
and five on t. His men move in the direction amnz, are re-entered in af and
borne in tz. The other player (D) has two men on z, five on n, three on h,
and five on f. They move in direction znma, are re-entered in zt and borne
on fa. This is the arrangement of our backgammon. Single men can be taken
and have to re-enter. The first player to bear all his men wins.
1 Allsingle-track games; the Délassemens du Mars, Marine, Jeu de !’oie have tracks of sixty-three
points, and only differ in the penalties or rewards attached to the arrival at particular points.
Mappemonde has a track of seventy-eight points; Guerre and Poule de Henri IV, eighty-three points,
and Hymen, ninety points.
2 The Persian board has a wider bar than the European board. This is to provide a space into
which the dice must be thrown.
5356 I
114 RACE-GAMES [CH. 6
Nard is the Persian word for a wooden block or cylinder, and shir means
‘lion’, the names referring to the two shapes of the men with which the
game was played, the simpler type being small plain cylinders of wood, and
the more elaborate, cylinders crowned with lions’ heads. Men of both types
were also used in the ancient Egyptian race-games. Popular etymology
associated nardshir with Ardshir (Artaxerxes), the founder of the Sassanian
dynasty (A.D. 226-41) and ascribed the invention of the game to this monarch.

FIG. §3. BOARD AND NOTATIONS FOR TABLES

The oldest mention of nardshir may be that in the Babylonian Talmud

(Ketuboth, 615) which was compiled between a.p. 300 and 500, though
additions continued to be made for another five hundred years. The passage
in the Talmud refers to playing with isqundari, ‘small dogs’, and nardshir. I
think there can be no doubt that the isqundari were the men with which
nardshir was played, which are called kilab, ‘dogs’, in the older Arabic men-
ions of nard, though Jewish commentators guessed otherwise; clearly they
had little knowledge of nardshir, for some explained nardshir as chess (which
it certainly was not), others as the same as the Arabic nard, and others as the
Greek kubeia.
Nard is also mentioned in the Middle Persian romance, Chatrang-namak,
probably written between 650 and 850, which treats of the introduction of
chess into Persia, the same story which is told by Firdawsi in the Shahnama.
In the earlier work, Buzurjmihr invents New-Artakhshir to confound the
Indian king.
As a result of the conquest of Persia, 638-51, the Arabs learnt the game of
nard, and an elaborate set of men, made of coral and turquoise, which belonged
to Khusraw (II) Parwiz (590-628), passed into the possession of the Muslim
caliphs (ath-Tha‘alibi). The victorious Muslims were playing nard before the
end of the seventh century, and it is recorded that abi Huraira (died 676-7),
one of the Prophet’s companions, refused to greet a group of Muslims who
were playing nard. The tabi‘ ash-Sha‘bi (died 722-3), of Persian descent, is
said to have played nard for a stake in the open street, though he covered his
head in a shawl while doing so, lest passers-by should recognize him. The
6.1.2] RACE-GAMES II§
reason for this was that the game was already suspect as under condemnation
in the Qur’an. The legal position was finally settled in the eighth century
when the founders of the four schools of Muslim jurisprudence concurred
in condemning nard as hardm, i.e. strictly forbidden to all Muslims, and in
attaching legal disabilities to players of nard. This decision did not prevent
the playing of nard at the courts of Hariin-ar-Rashid and his family, nor
hinder al-‘Adli (c. 850) from writing a book on the game. The caliph al-
Ma’min (died 833) preferred nard to chess because, if he lost, he could put
the blame on the dice and not on his own play.
There are several references to nard in Persian and Arabic works, dealing .
with the popular opinion as to its history and symbolism or the legality of
playing it. Al-Ya‘qibi (ante 872) makes nard an Indian invention to exemplify
man’s dependence on chance and destiny. The board stands for the year; it
has twenty-four points for the hours of a day: it is divided into two halves,
each of twelve points for the months in a year. The thirty men (kilab) stand
for the days in a month. The two dice are day and night, and the sum of
opposite faces of the die is seven for the days of the week. This symbolism
was known to the Byzantine Greeks (p. 27) and is developed at greater length
in MS. Brit. Mus., Arab. Add. 7515 (ff. 5b-6b).! A quatrain, composed
impromptu by Araqi (end 15th c.) to calm the dangerous anger of his patron,
the Seljuk king Tughanshah, whose temper had given way when he threw
two ones instead of two sixes at a critical point in a game of nard, is translated
thus by E. G. Browne (Literary History of Persia, 1909, p. 39):

Reprove not Fortune with discourteous tricks

If by the king, desiring double six,
Two ones were thrown; for whomsoe’er he calls
Face to the earth before him prostrate falls.

Nard has spread from Persia northwards to Georgia where Hyde (p. 3)
says it is called nardi, and to the Kalmucks of central Asia and the steppes
north of Astrakhan, where P. S. Pallas (i. 157) says that it is called narr. It
is also played by the Muhammadans of the Deccan, and tukhta-e-nard is
included by Herklots (App. vii) in his list of games played there. It does
not seem to have spread to the Hindus who had their own well-established
race-games. Nevertheless, it is certain that the older nard must have reached
northern India during the Sassanian rule in Persia. The Persian story in the
Chatrang-namak, if it stood alone, might be dismissed as fable, as indeed its
literary embellishments must be, but it is supported by Chinese evidence
that shwan-liu, ‘double-sixes’, was introduced into China from India in the
seventh century A.D.
6.1.3. China: Shwan-liu; Korea: Ssang-ryouk; Japan: Sunoroku or
Sugoroku. All these terms mean “double-sixes’ (Culin, a, 79-81; b, 499-504;
1 This manuscript has been edited and translated into Spanish by F. M. Pareia Casaiias (Libro
del Ajedrez, 1935), in which the game of nard is treated in i. 10-14.
 116 RACE-GAMES [CH. 6
Himly, a, 478-81). The Chinese board, cung ho inner table, ei kia; outer
table, wai hua (Culin, in T’oung Pao, ix, Oct. 1898) has a wider bar than the
European board, but less wide than the Persian board. The Korean board is
similar, but the men are tall pins, about 2} ins. in height, and one side is
coloured red. The Japanese board is usually made in the shape of a box and
omits the bar. All these games are played in the same way as nard, but in
Korea it is usual to play two men when doublets are thrown.
~~ Sunoroku is also used in Japan in the sense of ‘race-game’ and is used for all
1 ) kinds of race-games which are invented annually. Prof. Tsuboi told me in
‘) 1912 that the backgammon game was dropping out of use in Japan, but that
( a club of some thirty or forty members had been founded in Tokyo about
\\ I9to to revive the game.
6.1.4. Malaya: Main tabal (Skeat, 487; R. J. Wilkinson, 57). Two varie-
ties of the game are played, usually by women. The board from Johore
which Culin (b, 502) reproduces, contains two walled-in spaces instead of the
bar, in which the players throw their dice. The name, tabal, suggests that
the game was learnt from European traders.
6.1.5. Celebes: Pataballang (Kaudern, 310). According to Matthes only
played by the orang Macassar. A wooden board is used.
6.1.6. Celebes, Bugis tribe: Adangang; Macassar tribe: Padagang
(Kaudern, 310). The more common game in the coastal regions. Played on a
board set out on the ground. No description is given, and it is possible that
both Kaudern and Matthes have used the term backgammon loosely for
other race-games.
6.1.7. Siam: Len sake or saka (Capt. Low, 381; Culin, 6, sor). Played
on a board resembling that used in Johore (6.1.4). Two dice are thrown from
a special kind of dice-box with a hole on one side through which the dice are
dropped in the box from a hollow cylinder. Two players, each with sixteen
men instead of the usual fifteen, which are entered on the board by throws
of the dice. Neither description is very clear. Low says: ‘The men are not
placed in the box at the outset, but are kept in a heap in the chequers of
the players which are to the left hand of each. The pieces are filled into the
respective chequers according to the casts of the dice and they range to the
right when the whole numbers have been filled in. After this they may take
up any uncovered counters of the adversary which generally terminates the
game. The chief aim is to prevent a piece being uncovered.’ Culin says: ‘The
pieces are entered in the right-hand side of the board opposite the player,
and are moved round as in our game, to the side directly opposite, where
they are thrown off. A player does not take the opponent’s pieces.’
6.1.8. Asia Minor, Egypt: Tawula (Falkener, 256; M. S. Kadri). Two
players, each with fifteen men. Played on the backgammon board. C places
two men in a and D two in n; these cannot be moved until all the remaining
men have been entered. C enters in bg and D in ot, and both play in the
direction amnz. Neither player can double any men in his first two tables,
6.1.8] RACE-GAMES 117
except C on m and D on z. Blots may be taken and have to be re-entered
before the player makes another move and, if a blot is taken from the oppo-
nent’s first table, a point in that table must be left empty so that the opponent
has a chance of re-entry. The first player to bear all his men wins. If a player
has borne one man and has all his other men ready to be borne, while his
opponent has one man awaiting re-entry and the rest of his men in his entry
table, the first player wins capote, counting seven games. The winner then
places one man in e and one in f before tossing for the move in the next game.

2. THE EUROPEAN GAMES OF TABLES!

The European games of tables are derived from two sources which can be
identified by the number of dice used in playing them: one is the Roman
alea or tabula played with three dice, which survived the fall of the Roman
Empire and was still played in Italy and Spain when contact was made with
the Muslim world; the other is the Muslim nard which was played with two
dice. Both varieties had reached France in the course of the eleventh century
and were played by the crusaders of the First Crusade. The earliest mention
that I have found for Germany occurs in the will of Count Sibotos of Neuen-
berg (c. 1180. Monumenta Boica, vii. 502) in which he left four scahzabel and
four wurfzabel and ivory pieces (lapides) pertaining to these games. The game
may have been a century later in reaching Iceland.
The following terms were used generically in the medieval European
languages for any game played on the backgammon board: ML. ludus tabu-
larum, tabulae; It. tavole; Sp. tablas; Port. taboas; Prov. and Fr. dialects, taules;
MEF. tables; ME. tavel, tables; MHG. wurfzabel; MDu. worptafel, werptafel;
MDa. wartabl; MSw. vartafvel; Ger. brettspiel; Sw. bradspil; Cz. vrhcaby; Pol.
warcaby.
It will be noticed that the Latin, Romanic, and English terms are all plural,
and, following the analogy of the terms chess and merels, we should expect
that their singulars were the names of the tablemen with which the games
were played. As may be seen from many museum pieces, the tableman was
a flat circular disk, the same shape as our modern draughtsman, but rather
larger.? That the tableman was called tabula in Medieval Latin and the derived
Romanic languages is evidenced both by the BS. and CB. problem collec-
tions, and by a tenth-century poem, Versus Acbranni de ludo tabularum (MS.
Berne, 299, ff. 29b-30a), where, however, the game is not tables but a well-
known mathematical puzzle (Boy’s Own Book, 699) with counters. In MF.
the tableman is always table or taule, and in the Norse languages tabl and
tafl are regularly used as the name of the tableman.
The medieval board for tables obtained its name from the tableman; ML.

I For the medieval games see my article on them in Medium Aevum, x (June 1941), $7—-69.
2 The medieval backgammon board was double the size of the chessboard and the tablemen
were too large for use on the cells of the chessboards. Museum pieces are usually labelled “draughts-
men’, but this is wrong; they were tablemen only.
118 RACE-GAMES [CH. 6
tabularium; It. tavoliere; Sp. tablero; Prov. taulier; MF. tablier; ME. tabler. The
half-board is casa or quadra in Alf., pagina in K., casa in the Italian BS. MS.,
and table in England, whence arose the English custom of calling the whole
board a ‘pair of tables’ which already occurs in an inventory of the goods of
Sir John Marmadux made in 1314 (Reg. Pal. Dunelm., Rolls Ser., ii. 673-6)
including ‘i. par tabularium and i. par de tabulis’. Joan Steven’s will of 1459
(Bury Wills, 180) has ‘vnum par de tabulis cum chesmen et tablemenys’. A
player’s bearing table is his domus in BS. and CB. The partition between the
outer and inner tables, our bar, is barra in Alf. Each table contains six points
(Lat. punctus or punctum) along its opposite edges, and the points are named
ace-point, &c. The dice-point (le si in CB. 6) is frequently named caput, It.
capocchia, and MF. teste.
There are two medieval treatises which deal at some length with tables.
The Alfonso MS. describes fifteen games played in Spain in 1283, and K.
describes eight varieties played c. 1300 in England. The two collections of
problems, BS. (ante 1300) and CB. (1400-50), contain in all eighty-four
problems of tables, seven varieties being named. These are of two kinds,
optative games (Iudi optativi) in which the players choose their throws instead
of throwing the dice, and problems in which the dice are used. The latter are
really mathematical exercises in probability, and their difficulty is enhanced
by conditions governing the use of the dice. Players had first to agree whether
to play ad postam or ad invitum, terms of which no explanation is given,
presumably because they were well understood. In several problems ad
invitum is prescribed, in some it cannot be refused, in others the solution
depends on a right choice, thus CB. 66, ‘si sit ad postam, elige albas, si ad
invitum nigras.’ K. also includes six jupertiae, ‘jeopardies’, with their solutions,
none of which occurs in BS. or CB.
The names of the various throws of a single die are in BS. and CB. as, du
or deus, trei, quater, cinc, si or sis. Equal throws of two dice (doublets) are
ambas or amesas, du du, tern or ternes, quaterne or quaternes, quine or quines, and
sine or sines. Equal throws of three dice are in Alf., amesas el terz, du du el terz,
&c., and in BS. and CB. are buf or buffe de as, du, &c., and in one CB. problem
(MS. Florence, xix. 7, 37, f. 207b) are 2,2,2 dexeot, 6,6,6 sexeot. But three dice
were not always available, in which case different ways of obtaining the
third throw were adopted. Sometimes the third throw was assumed to be a 6,
sometimes the player or his opponent chose the throw. But more common
were the customs of majoret, by which the higher throw of two dice was
duplicated, and minoret,! by which the lower throw was duplicated. In the
older games doublets were treated no differently from unequal throws and
gave no extra throw. Ifa player could not use a throw, it was generally lost,
but in some games could be used by the opponent. If the players agreed to
play ad fallum, MF. 4 Ia faille, inability to use any throw involved the imme-
diate loss of the game.
Not to be confused with the game of the same name.
6.1.8] RACE-GAMES II9

The popularity of tables in the Middle Ages was very great. Chess and
tables are regularly mentioned together as the typical games of the upper
classes of the twelfth and following centuries, and many quotations establish-
ing this will be found in HC. (428-42, 447-51). Chess is named first as the
more difficult game and this is definitely implied in the Chanson de Roland
(c. 1165, ll. 110-12):
Sur palles blancs siedent cil chevalier,
As tables juent pur els esbaneier,
E as eschas li plus saive e li veill.
From the upper classes tables spread to all other orders of society. Innkeepers
provided boards and men for their customers, and doubtless attracted custo-
mers thereby.! It was a charge made against the clergy in Provence (B.P.
Bartsch, Chrestomathie provengale, 374-7) that they haunted taverns to play at
tables and hazard.
Against this popularity of tables the Church waged a long and losing war.
In 1254 St. Louis IX of France not only forbade his court officials to play
tables but extended the prohibition to all his subjects (Marténe and Durand,
Thesaurus novus Anecdotorum, Paris 1717, i. 437), and some 150 years later,
the Archbishop of Tournai was active in prosecuting players of tables (Gode-
froi, sv. table, tableour). When the schools of Bologna decided that the
ecclesiastical canons did not apply to chess, tables was still classed with the
inhonesti ludi and it was not until the end of the fifteenth century that these
attempts to suppress the game were abandoned. By that time the custumals
of towns in Italy, France, Spain, Switzerland, and Holland were specifically
excepting tables from municipal censure, provided the stakes were small, and
only the universities and the regulations for apprentices preserved the stricter
position. The last injunctions against tables in England are to be found in the
Elizabethan Canons of 1571 and the consequent articles for various dioceses
(Kennedy, 1. Ixxxviii; m1. 30, 64; mI. 340). In the eighteenth century back-
gammon was a favourite game with the country clergy in England.
Lists of games of tables are given in several works of the sixteenth century:
for Italy in Cardan’s De Ludo Aleae (1524-30; see Hyde, 40) and in Citolini’s
Tipocosmia (1561, p. 484; see Fiske, 223); for France in Rabelais (1. xxii. 1534);
and for Germany in Johann Agricola’s work on proverbs (1529; cf. Fiske,
255). The Italian games given by Cardan are sbaraino sbaraia and sbaraiono
(Cit., sbaraglio, sbaraglino), toccadiglio parvum and magnum, canis Martius (Cit.
camarzo), minoretto major and minor, and jacina. Rabelais’s French games are
tables, nicque nocque, lourche, renette, barignin, trictrac, toutes tables, and tables
rabatués. The German games are biif or buf (three kinds), frawenspiel, verkeret
lang, and kurz, lorzen, and dickedack. The Keurboek of Delft, Holland (c. 1525),
1 The name ‘Chequers’ for inns and the signboard, if not derived from the armorial bearings
of the lord of the manor, may have arisen in this way.
2 Later known as kurz and lang puf and gegenpuf. Grimm’s etymology ofpufas a native German
word is clearly wrong: buf and bufa were the names of an Italian and Spanish game before the
game was played in Germany.
120 RACE-GAMES [CH. 6
permits play at verkeer and ticktack. Verkeer is also mentioned in the expenses
of the Dukes of Burgundy, 1379-80 (La Borde, Les Ducs de Bourgoyne, 1851,
ii. 287), ‘4383. Item vii Octobris, dominae personaliter ludenti cum ludo
dicto ter quarten. 4384. Aleae, viz. tverkeerde.’! Six games are mentioned in
English works of the sixteenth century, ticktack, queen’s game, doublet, fails,
lurch, and most frequently, Irish. The first reference to the last game occurs
in Barclay’s Shyp ofFolye (1509, p. 14): ‘Thoughe one knowe but the yrische
game, Yet wolde he haue a gentyllmannys name.’ The use of these national
varieties of tables led gradually to the disuse of the generic term, tables.
There is naturally a general likeness in all the major games of tables, and in
the following descriptions I shall assume (unless otherwise stated) that they
are played like backgammon, the men moving along a given track and being
borne from the home table, the piled men immune from capture and the
single man or blot (K. homo vagans, BS. solus, CB. nudus, solus, or unicus)
vulnerable and to be re-entered after capture. The special features alone will
be given; these concern the use of the dice, the use of the whole board or a
portion only, the entry table (E), the direction of move (M) and bearing
table (B), the restrictions on piling in certain parts of the board, and the
values attached to particular ways of winning. I use the notation of the points
shown in Fig. 53, and follow the example of the French Le grand Trictrac,
Paris 1756, in calling the players Chloris (C) and Damon (D).
GAMES OF ENTRY, PILING, UNPILING AND BEARING ONLY

6.2.1. Spain: Los doze canes (or hermanos) (Alf. 736). Two dice; a throw
that cannot be used is lost. Each player has twelve men which have to be
doubled in af. Blots can be taken and must re-enter before another man is
entered. Played by boys.
6.2.2. Spain: Doblet (Alf. 74a). Three dice; a throw that a player cannot
use is played by his opponent. C has two men on each point of af, D two on
each point of zt which have to be borne.
6.2.3. England (14th c.): Paumecary (K. 1586). Two dice; doublets give
a second throw. Played by two persons or sides of two or three persons
when all those of one side and then those on the other side play in succession.
Play is confined to af. Men can be piled and blots can be taken and then must
re-enter. When either side has entered all its men, it proceeds to bear them,
and when it has borne all its men, it proceeds to help bear the other side’s
men, smacking the losers’ hands once for each of their men so borne.
6.2.4. England (16th and 17th cc.): Catch-dolt, ketch-dolt (Cotton,
R. Holme). Two dice. C has to enter fifteen men in af, D fifteen in zt. When
entering, a player must draw a man which his opponent has entered on the
Opposite point, and only use one in hand when there is none to draw. If he
' Verkeer is usually derived from the Ger. verkehren, ‘to return’, because the men return in
some way. But the only form of return in verkeer is that common to all the major games of
tables, and the Dutch forms quoted above and the fact that the game is identical with the Icelandic
kotra make this derivation impossible.
6.2.4] RACE-GAMES 121

fails to do so, he is dolted and loses outright. When a player has entered fifteen
men in his table, he proceeds to bear them, and the first player to bear all his
men wins.
6.2.5. Tudor England: Doublets (Cotton, R. Holme). Two dice; doub-
lets give a second throw; any throw that a player cannot use is played by his
opponent. C has two men ona, b, c and three on d, e, f; D two on z, y, x,
and three on w, u, t. The players then bear all the men they have on any
point in excess of one, and then the remaining men.
Cotgrave says that this game resembles, and Hyde that it is the same game
as, queen's game or Fr. renette.
6.2.6. France: Tables (later, Dames) rabbatuées or avalées (Rabelais;
Jeu de Trictrac, 1699; Encyc. Méth., 1792). An extended form of 6.2.5. in
which, after a player has borne all his men, he has to re-enter them and
restore the original arrangement.
6.2.7. Iceland (18th c.): Ofanfelling, ofanfellingartafl (Isl. Gatur, 315).
Two dice; doublets give a second throw; a throw that cannot be used is lost.
C has two men on each point of af, D on each point of zt. Each player proceeds
first to unpile his men, next to repile them as before, and finally to bear them.
GAMES OF MOVEMENT WITH FEWER THAN FIFTEEN MEN A SIDE
6.2.8. Iceland: A6é elta stelpur, ‘chasing wenches’ (Isl. Gatur, 315). Two
dice, but the only throws that are used are any doublet, 6 and 1; doublets
give a second throw; a throw of 1 moves one man, of 6 or any doublets
except 6, 6, two men, and 6, 6, four men. C has one man on each point of af,
D one on each point of ns. Both sides move in direction amnz and continue
to circulate round the board until one player has lost all his men. No point
can be piled and if a throw brings a man to a point on which the player has
already a man, it is placed on the first vacant point beyond. Captures are not
re-entered. If a player is reduced to a single man (hornaskella, ‘corner rattler’)
he jumps from corner to corner and cannot be taken as long as he has an
opponent man in front of him.
6.2.9. France: Tourne-case (Jeu de Trictrac, 1699; Encyc. méth.). Two
dice, but doublets only count as a single throw (6, 6 as 6, &c.). Each player
has three men. C enters in af and uses af and gm only, D enters in zt and uses
zt and sn only. A player’s men must always be in the order of entry; no man
can pass over any in front of it, nor be piled on any point except his coin de
repos (C m, D n). If a player plays one of his men to a point immediately
Opposite one occupied by an enemy man, he takes it and it has to re-enter.
The first player to pile his three men on his coin de repos wins, and if he does
so before his opponent has entered any on his coin de repos, he wins double.
GAMES OF MOVEMENT WITH FIFTEEN MEN VARIOUSLY ARRANGED
ON THE BOARD
6.2.10. Spain: Las quinze tablas (Alf. 73a). Three dice C with E af;
M amnz; B tz; D with E zt; M znma; B fa. C has two men on each point
122 RACE-GAMES [CH. 6
from a to g and one on h, D has two on each point from z to s and one on r.
The game may be played with any smaller number of men, thus, las seys
tablas (Alf. 732) is played with sides of six men.
6.2.11. Spain: Todas tablas (Alf. 77b); France: Toutes tables; Italy:
Tavole reale; England: Irish (Cotton). Two dice; doublets do not give a
second throw. C has E af; M amnz; B tz, and two men on a, three on r, and
five on m and t; D has E zt; M znma; B fa, and two men on z, three on h
and five on n and f. This is the arrangement in backgammon and in most of
the Asiatic games.
6.2.12. England: Backgammon; Scotland: Gammon; Spain: Tablas
reales; Italy: tavole reale (Cotton; Hoyle, &c.). The modern form of 6.2.11,
invented in England in the early years of the seventeenth century, from which
it differs only in that doublets are played twice and wins are graduated as
follows:
Backgammon, counting 3 (and, according to Cotton, 4 if the last men
are borne by doublets), when the winner bears all his men before his oppo-
nent has carried all his men to his home or bearing table.
Gammon, counting 2, when the winner has borne all his men before his
opponent has borne any.
Hit, counting 1, when the loser has borne some of his men. The first men-
tion of this game occurs in J. Howell, Epistolae Ho-elianae (1650, ii. 105):
‘Though you have learnt to play at Baggamon, you must not forget Irish,
which is a more serious and solid game.’
In modern times changes in the use of the dice have been introduced. In
Scotland some players, when doublets are thrown, also play the reverse
doublets, and allow another throw if all doublets and reverse doublets have
been played. Other players use acey-deucy (a throw of 2, 1), playing first 2, 1
in either order, then 6, 6 and 1, 1, and if all these have been played, throwing
again (E. C. Boyden). In Italy wins are not graduated and both doublets
and reverse doublets are played, though some players do not play the reverse
doublets when bearing, and doublets do not give another throw. If 2, 1 is
thrown, the player can either play 2, 1 or any doublet that he likes, but this
does not give another throw. If a player cannot use a throw, his opponent
may use it.
The English and Scotch names are probably adopted from the special
victories (as limpole and lurch took the place of Emperador (see 6.2.25). After-
game and back-game are wrongly defined in OED.; both mean play in which
the player hangs back with his men instead of pushing them forward as fast
as he can (Hoyle, 42).
6.2.13. Italy: Barail, Desbaraill (BS., CB.); Sbaraglio (Cardan). Three
dice. Both players have E af; M amnz; B tz. C has one man on k and fourteen
on 1; D has fifteen on m.
6.2.14. Italy: Barail, Sharaione (Cardan). Only differs from 6.2.13 in
that throws are played twice.
6.2.15] RACE-GAMES 123

6.2.15. Italy: Barail, Sbaraglino (Bartinello, II nobile et dilettevol Givoco

del Sbaraglino, 1604); England: Baralie (K. 159a). Only differs from 6.2.13
in that it is played with two dice and an invariable throw of 6 instead of a
third die. In Italy, a player who bears all his men before his opponent has
borne one wins marcio.
6.2.16. Spain: Cab e equinal (Alf. 77a). So named because all the men
are arranged in one table. Three dice. For both players, E af; M amnz; B tz.
C has fifteen men on g, D fifteen on h.
6.2.17. Spain: Cab e equinal (Alf. 77a). Two dice and an invariable
throw of 6 instead of a third die. Only differs from 6.2.16 in the arrange-
ment. C has fourteen men on g and one on t. D has fourteen on e and one
on d.
6.2.18. Spain: Fallas (Alf. 746); France: Faille; England: Fails (K. 159a).
Three dice, but if only two are available, minoret. If a player is unable to
play any throw, he loses outright. C has E af; M amnz; B tz and two men
on a and thirteen on t. D has E zt; M znma; B fa and two men on z and
thirteen on f.
6.2.19. Italy (where it was the leading variety in the Romagna in 1600);
Minoret (BS.; CB.; Bartinello, op. cit.). I have found no description of
this game, but two varieties were played; major or Iungo and minor or curto.
Three problems in BS. and CB. suggest that the men were arranged as in
O:2 ellos
6.2.20. England (14th c.): Imperial (K. 1594). Three dice. C has E af;
M amnz; B tz and five men on f, g, and h. D has E zt; M znma; B fa and
five men on t, s, and p.
6.2.21. Spain: Laquet (Alf. 78a, where it is said to be a recently invented
game). Two dice; if a player cannot use a throw, his opponent can use it.
No captures are permitted in this game. Both players have M amnz and B tz.
C has one man on a and fourteen on z. D has one on m, two on t, and three
on u, w, x, and y.
6.2.22. England: name unrecorded but Strutt calls it Myles (K. 1596).
Two dice and an mvariable throw of 6 instead of the third die. Both players
have E af; M amnz; B tz. C, who plays first, has five men on m, four on r
and s, two on x; D has three on n, 0, p, q, and r.
6.2.23. England: Provincial (K. 159a). A variety of 6.2.20. Three dice.
C has E af; M amnz; B tz and eight men on fand seven on g. D has E zt;
M znma; B fa and eight men on t and seven on s.
6.2.24. Spain: El seys, dos e as (Alf. 75a). Three dice. Both players have
E af; M amnz; B tz. C has three men ona, four on b, and eight on f (whence
the name of the game). D has five men on each of c, d, and e. Captured men
are re-entered in the table diagonally opposite to the one in which they are
taken.
Fiske (pp. 88, 300) identifies this game with the English sice-ace, but this
is wrong; sice-ace was played with dice but was not a board-game.
124 RACE-GAMES [CH. 6
GAMES OF MOVEMENT—starting from the arrangement C fifteen men on a,
D
fifteen on z; C with E af, M amnz; B tz, D with E zt, M znma;B fa.
6.2.25. Spain: Emperador (Alf. 756); England: Ludus Anglicorum
(K. 1582); Italy: Testa (BS., CB.); France: Tieste (BS.). Three dice. In
England if only two dice are available, an invariable throw of 6, for the
third die, or the player or his opponent chooses the third throw. Neither
player can re-enter a man on a point already occupied by one of his own
men or pile men in his half of the board. Wins are graduated as follows:
Barata (Alf.). The player has doubled men on six consecutive points.
Limpolding (K.). One player has doubled men on the first five points on
the opposite side of the board and his opponent has eight men on the ace-
point of his bearing table, one man on the first point of his inner table and
on each point of his entry table, and one man in hand to be re-entered.
Lurching (K.). Only differs from limpolding in that the opponent has
fewer than eight men on the ace-point of his bearing table.
Apparently, as in our backgammon, the special names of these victories
were also used for the game itself: a problem which in most BS. MSS. is
of testa is in a French version (Montpellier, H. 279, 935) said to be of la linpole.
Sixteenth-century works often refer to a game of tables called lurch, Fr.
lourche, Ger. lorzen, lurtschen, lurtsch, though none describes it.
6.2.26. Italy, France: Imperial (BS., CB.). Two dice with an invariable
throw of 6 for the third die. Only known from four problems, but there is
nothing in them that is inconsistent with the game being 6.2.25.
6.2.27. Spain: El medio Emperador (Alf. 76a); England: Ludus Lom-
bardorum (K. 1592). The same as 6.2.25 but confined to the two tables af
and tz.
6.2.28. France: Garanguet (French game manuals). Three dice. If all give
the same throw, it is played triply; if two give the same throw and their sum
exceeds the throw of the third die, the equal throws are played twice.
6.2.29. France: Trictrac, Petit trictrac (Divert. Inn.); England: Tick-
tack (Cotton); Germany: Dickedack. Invented in France about 1500, and
the parent of a number of games in which the race character is obscured,
since the games are not won by bearing but by scoring points for the different
possibilities of move given by the throws of the dice. The following technical
terms are used:
CHD)
Points: Talon ; a Zz
Coin, coin de repos : 5 fel aD
Tables: Petit jan. : : Sowa, eoZE
Grand jan . : : . gm sn
Jan de retour : : , ta

Two dice. No captures can be made (cf. 6.2.21). In each turn of play the
player throws the dice and looks to see in what ways the throw might be
used. If he can take a blot, he scores 1; if he can post two men on his coin, he
6.2.29] RACE-GAMES 125

scores 2; if he can double men on each point of his grand jan, he scores 2. He
then enters up his total score so obtained. If he has omitted to register any
possible score, his opponent takes this score by a ‘why not?’ Finally, the
player makes his move.
Cotton says that some English players played with toots (to fill petit jan),
boveries (to post one man in m and one in n), and flyers (to advance a man to
jan de retour before his opponent has left his petit jan).
6.2.30. France: Grand trictrac (Divert. Inn., Grand Trictrac, 1756). An
improvement of 6.2.29 which it replaced in the reign of Louis XIV. Now the
leading variety of tables played in France.
Two dice. Moves are only made by the players in their own half of the
board, though scores can be made by the ability to make a hit in the oppo-
nent’s half; thus, if C can hit a blot in an, he scores 2 for each way of so
doing, or 4, if by doublets.
A player can only fill his coin by entering two men together on it, scoring 4
(6 by doublets), and can only fill his opponent’s coin after he has filled his
own. If, however, a player could fill his opponent’s coin if his own had not
been empty, he fills his own instead par puissance, making the same score.
Scores are also made as follows:
Jan de trois coups or six tables (4). The third move gives a man on each of
the points b-g (y-s).
Jan de deux tables (4, by doublets 6). Only two men have been played from
the talon, and the throw would enable the player to occupy both m and n.
If, however, the opponent has already occupied his coin, it is contre-jan de deux
tables, and the opponent takes the score.
Petit jan and grand jan (4, by doublets 6 for each way possible). The player
has piled men on each point of the jan.
Jan de mezeas (4, by doublets 6). The player, having two men in his coin
while his opponent’s coin is empty, throws a 1 or 1, 1. If, however, the oppo-
nent’s coin is already filled, it is contre-jan de mezeas and the opponent takes
the score.
For every move which preserves petit or grand jan, the player scores 4 (by
doublets 6) even if the position is merely preserved by inability to play a
throw in whole or in part. The opponent scores 2 for each man which cannot
be played by a throw.
Before 1700, three other jans were recognized:
Jan de rencontre (2). Scored by the second player if his first throw repeats
the first player’s throw.
Jan qui ne peut (2). A throw, taken singly or combined, hits points on
which the opponent has piled men and no other move is possible.
Margot la fendue (2, by doublets 4). A throw brings a man to an empty
point between two enemy men.
The player must mark his score before making his move. If he omits to
count any score or claims one wrongly, his opponent takes that score, sending
126 RACE-GAMES [CH. 6

the player to school. The game is for twelve trous, and twelve points make
one trou. The points are marked along the middle of the board and the trous
by pegs in holes on the outer edges of the board. If one player scores twelve
points before his opponent scores one, he marks two trous and may go away,
i.e. end that game and begin a new one, carrying forward any trous he may
have scored to the new game.
6.2.31. France: Grand trictrac; jeu de retour. Two dice. Differs from
6.2.30 in using the whole board. In addition to the above scores, the player
scores 4 (by doublets 6) when he can fill his jan de retour. The first player to
bear all his men scores 4 (6 if his last throw was doublets). If neither player
has then scored twelve trous, the score is carried forward to a new game.
Two men piled on the coin cannot be moved separately. An empty coin
can be used as an intermediate step when playing a combined throw of the
dice. Neither player can take his opponent’s coin or enter his petit or grand jan
so long as his opponent has a chance of filling them.
Before the middle of the eighteenth century, pile de misére scored 4 (by
doublets 6). This happened when one player had fifteen men piled on his
coin and his opponent had men piled on every point of his grand jan.
6.2.32. France: Trictrac 4 écrire (Grand Trictrac, 297). A method by
which three players can play 6.2.30 or 6.2.31.
6.2.33. France: Plein. Two dice; doublets are played twice, but in playing
from the talon no point can receive two men, so 3, 3 moves one man to k
and the other to d. C uses am only and D zn. A single man can be played to
a coin. The player wins who first fills his grand jan.
6.2.34. France: Toe (Jeu de Trictrac, 169 and other manuals). Two dice.
A variant of 6.2.31, differing only in the rules for scoring. Generally identi-
fied with the following game, but this is doubtful.
6.2.35. Italy: Toccateglio (Cardan); Germany: Tokkadille (Fiske, 259).
Italian manuals say that this game is of Spanish origin, but there is no Spanish
evidence that this is so. It was played in two ways, curto and lungo, but no
contemporary description is known. German manuals say that it is occasion-
ally played with sixteen men a side.
6.2.36. Sweden: Schuster (Fiske, 318). A women’s game. Two dice: a
variant of 6.2.30. Ifa player makes every point in her petit jan (Sw. schuster)
with three men on each of its last three points, it is schuster, scoring 2 and the
game is over. If neither player makes schuster, the game is played through the
whole board, and the first to bear all her men wins. As in 6.2.30, blots cannot
be taken.

GAMES—starting from the arrangement C fifteen on a, D fifteen on n; C with

E af; M amnz; B tz; D with Ens; M nzam; B gm. 6.2.37. Spain: Chaquete;
Catalonia and France: Jacquet; Italy: Giacchetto (Fiske, 193). Two dice;
doublets are played twice. One man (Fr. courier; It. corriere) must be played
first and no other man can be played until the courier has reached its bearing
6.2.37] RACE-GAMES 127

table. Only two points in the player’s entry table may be doubled at the
same time, and C can only have two men on m (D on z), otherwise any
point on the second half of the board may be piled. Men that fail to reach
their bearing table are called cochinnets. A player wins marcia (counting 2) if
he bears all his men before his opponent bears one, and marcia per punto if he
gets marcia without doubling men on six consecutive points.
Possibly the same game as Cardan’s jacina, but that was played with three
dice.
6.2.38. France: Jacquet de Versailles (Moulidars). This only differs from
6.2.37 in the treatment of doublets; 6, 6 counting as six 6’s, 5, 5 as five $’s, &c.
6.2.39. Holland: Verkeer; Germany: Verkehren; England: Verquere
(Seymour, 1754); France: Reverquur Revertier; Denmark: Forkoering;
Sweden: Svensk bradspel; Iceland: Kotra. Two dice; doublets are played
twice. Neither player may pile any point (except the last) on his first half of
the board. Men cannot be re-entered on any point that is already occupied
by one of his men. In Germany before starting a game the players must
agree whether to play with five blots (i.e. to play out five men before doub-
ling men) and whether to have more than five doubled points at a time. If
a player has men to re-enter and there are no points available, he is jahned or
junkered. In France and England there is no limit to the number of points
that may be doubled in a sequence. In England a player who is ‘johned’ (i.e.
has more than six men to re-enter) loses double.
In Sweden the higher throw must always be played first and throws may
not be combined to re-enter a man. A player may not double more than five
consecutive points until his opponent has borne all his men but one.
Wins are graduated in Sweden and Iceland. I describe the wins as for C.
The Swedish wins are: (1) bearing all men first; (2) enkelt kronspel, five men
on each of x, y, and z; (3) dobbelt kronspel, three men on each of u, w, x, y, Z;
(4) trappspel, seven men on z, five on y, three on x; (5) uppspel, fifteen men on
z; (6) jan, the opponent has more men to re-enter than points available; (7)
sprangjan, the opponent has no points on which to re-enter.
In Iceland when a player held six points in sequence his opponent formerly
lost his turn of move until some of these points were vacated. Now, he can
hit any of these points as though it was a blot. The Icelandic names for the
Swedish wins and their values are: (1) uttekt, 3; (2) stutti mukur, ‘little monk’,
7; (3) langi mukur, 5; (5) meistari, 13; (6) and (7) jan, 15. To which the Ice-
landic game adds (8) langi hryggur, ‘long back’, three on each of t, u, w, x,
y, Z; (9) half monk, scoring a monk by the throw of one die, the throw of the
other being discarded; (10) little meistari (not generally recognized), seven
on y, eight on z. All these scores are increased by 2, mar, if they are made
while the opponent has a man awaiting re-entry (Fiske, 316-29, 350-2).
6.2.40. Iceland: Piprjall (Fiske, 352). A variety of kotra (6.2.39), only
differing in the play of doublets. If 6, 6 is thrown, all the doublets are played
in order from 6, 6 to I, 1; if 5, 5 is thrown, 5, 5 and 6, 6 are played; if 4, 4
128 RACE-GAMES [CH. 6
or 3, 3 or 2, 2 is thrown all the smaller doublets are played in downward
order, including the throw; if 1, 1 is thrown, all the doublets in rising order
are played.
6.2.41. Sweden: Verkehren utan junker. This is 6.2.39 without the
junker feature. If sprangian occurs, the player can treat the point on which
he should enter as a blot.
GAMES WITHOUT INITIAL ARRANGEMENT. (a) Both players haveE af;
M amnz; B tz. 6.2.42. Italy: Buf, buffa (CB.), or Buffa cortesa (Alf. 785);
Spain: Pareia de entrada (Alf. 76b). Two dice (CB.); two or three dice
(Alf.); doublets are played twice and give a second throw. Throws that a
player cannot use are used by his opponent.
6.2.43. England: Paumecary (K. 158b). Two dice; doublets are played
twice and give a second throw. Played in the same way as 6.2.3.
6.2.44. Germany: Lange Puff; England: German or Russian back-
gammon (Fiske, 267-8). Two dice; doublets are played twice, then their
reverse doublets twice, and the player throws again. If this again gives
doublets, they and their reverse doublets are played but there is no third
throw (Germany). In England the first throw of doublets gives no second
throw. Reverse doublets are not played when bearing men. If deucy-ace
(2, 1) is thrown, the player can substitute any doublet he likes, but there is
no second throw. If a player cannot use a throw, his opponent uses it. This
is the modern form of 6.2.42.
GAMES WITHOUT INITIAL ARRANGEMENT. (b) C has E af; M amnz;
B tz; D has E mg; M mazn; B sn. 6.2.45. Spain, Italy: Buf de Baldriac
(Alf. 79a; CB. 73). Three dice. A throw that a player cannot use is lost.
GAMES WITHOUT INITIAL ARRANGEMENT. (c) C has E af; M amnz;
Btz; Dhas Ens; M nzam; B gm. 6.2.46. Spain: Los Romanos rencontrat
(Alf. 79b). Three dice. No man can be doubled in the player’s first half of
the board.
GAMES WITHOUT INITIAL ARRANGEMENT. (d) C hasE af; M amnz; B tz;
D has E zt; M znma; B fa. 6.2.47. Germany: Gegenpuf, Contrabuf,
Contrire puff (Fiske, 254-6); England: French or Double backgammon
(Boyden, 90). Two dice; the lower throw must be played first; doublets are
played twice, then their reverse doublets are played twice, and the player
throws again. If this again gives doublets they are played in the same way,
but there is no third throw. In Germany, the players first throw the dice to
determine who shall begin, and the winning throw is called general-pasch,
and whenever this throw occurs in the course of the game, it is played twice,
then the reverse numbers are played twice, but there is no second throw. No
man can be moved a second time until all the player’s men are entered.
Otherwise, throws can be played either singly or combined. A man after
taking a blot must remain on that point until the next throw.
6.2.48] RACE-GAMES 129

6.2.48. U.S.A. navy: Acey-deucey (Boyden, 93). A variety of 6.2.47 in

which acey-deucey takes the place of general-pasch. If 2, 1 is thrown, the
throw is first played, then any doublet and its reverse doublet that the player
chooses is played and the player throws again. When bearing men, the exact
throw must be used; a throw of 6, 6 when the side-point is empty, cannot
be used to bear from any other point and the throw is lost.
GAMES OF WHICH ONLY THE NAMES ARE KNOWN. 6.2.49. Italy: Bethe-
las, bathelas (CB. 72). Three dice.
6.2.50. Germany: Buf regal (Fiske, 158, 256).
6.2.51. ‘Italy: Canis martius (Cardan); Camarzo (Fiske, 171, 223).
6.2.52. Iceland: Danumannstafl, dandimannstafl (Isl. Gatur., 319; Fiske,
354).
3. GAMES ON SQUARE BOARDS
These would seem to be of Indian invention. In the oldest games the
board has an even number of cells but in all existing games the board has an
odd number. All boards contain a number of marked or crosscut cells which
include the central cell on odd-sided boards and the four central cells on
even-sided boards from which men are borne, the middle cells of the four
sides which serve as cells of entry, and occasionally the four corner cells.
Men occupying a crosscut cell are immune from capture. When two persons
play, they sit on opposite sides of the board and when four play, each sits at
a side of the board. Each player enters his men on the crosscut cell before
him! and moves anticlockwise round the border of the board and then clock-
wise round the inner cells until the men reach the central cell to be borne.
As a rule moves are given by the throws of four cowries and have as value
the number of mouths uppermost; if none are uppermost, the throw counts
as 8. Certain throws carry a ‘grace’ which
increases the value by one, and this one Scie ctl en
may be played (separately from the Gees

Le
number thrown) to enter or bear a man. igs ik: Gl clos emaa
In some games the throw may be divided > PW, Rey
to move two men instead of one. tia DX a
6.3.1. Ancient India: Ashtapada, ‘eight-
square’ (HC. 33-42).
The board contains 8X8 cells. It was
on this board that the inventor of chess i sf
arranged his game, and the Indian chess- Slee eae |eat
board to this day preserves the crosscut FIG. $4. ASHTAPADA
cells of the ashtapada board although they
play no part in chess, and they have even survived the chequering of the
board. The complete set of crosscut cells is shown in Fig. 54 but modern
! This applies also to cruciform boards, but in some games the men are placed on more ad-
vanced points of the tracks.
6356 K
130 RACE-GAMES [CH. 6
boards sometimes omit some of them. Dasapada, ‘ten-square’, was probably
a similar game on the 10X10 board. The game on a 66 board, depicted
in the gambling scene, Chitupada Sila, on the coping of the Stupa of
Bharhut, a monument assigned to the fourth century a.p. (A. Cunningham,
Stupa of Bharhut, 1879, pl. xlv), is probably a similar game.
The invention of chess did not interfere with the practice of the older
race-game and both games continued to be played on the same board. But
the term ashtapada passed out of use and in southern India both games were
known as chaturanga, ‘chess’. In the later literature we can only tell which
game is intended from the context; if the men are called sari, we know that
the race-game is meant. Finally, chaturanga was only used for the race-game,
and it became necessary to invent a new name of chess (HC. 61). We only
know how ashtapada was played from existing games.
6.3.2. Southern India, Ceylon: Saturankam, ‘chess’, Siga (Parker, 605-
7). Played by two, three, or four persons on a board of 9x9 cells with five

FIG. §§5. SATURANKAM

crosscut cells. The central cell is known as tachi and the other crosscut cells
are katti in Ceylon, and the plain cells are kddu in Tamil and gaeta in Ceylon.
Each player has two men (topparei in Ceylon). Two large four-sided dice
(kemadi) made of brass and with rounded edges are used, the values (1, 3, 4, 6)
are indicated by small holes in the faces. When throwing the dice, the player
first rolls them between his palms and then along the ground. Throws can
be combined to move a single man, or the total may be divided in any con-
venient way to move two men. A throw of doublets gives another throw.
Every throw must be used; a player cannot refuse a throw. The game begins
by each player stationing his men on his kafti. Single men on a kédu can be
taken and must be re-entered but this requires a throw of 1, 1. When a man
is four cells or less from its tachi, it can only be borne by a throw of the
appropriate doublet.
6.3.3] RACE-GAMES 131
6.3.3. Bengal: Ashta-kashte (Falkener, 266). Three or four persons play
on a board of 7x7 cells (Fig. 56). Each player has four men which he has to
enter on the crosscut cell on the edge nearest
to him. Men on crosscut cells cannot be
taken, with one exception: a player re-enter-
ing a man on his entry cell can do so even if
an opposing man occupies it. Single men on
plain cells can be taken, and two men piled
on these cells can be taken by playing two
men together to their cell. Moves are given
by the throw of four cowries, with the
values given above, and a capture or
a throw of 4 or 8 gives a grace and another
throw. A player can refuse to use any FIG. §6. ASHTA-KASHTE
throw.!
6.3.4. Southern India, Ceylon: Siga (Parker, 607-8); Sadurangam
(MVB., i. C. 5708a). Two, three, or four persons, each with two men, play
this game on a board of 5 ¢5 cells (Fig. 57). When
the board is drawn on the ground small sticks are
used for men, each player having two. Moves are
given by the throws of four cowries with the usual
values and graces. A grace is necessary to enter or re-
enter a man. A throw of 1 or 8 or a capture gives
another throw. Each throw moves one man except
when bearing, when a throw can be divided to bear AS
TO er . FIG. 57. SADU-
6.3.5. Penang, Malaya: Main pachih (O. T. eteren
Dussak, 69). Played by four persons on the board of
Fig. 55, each having two men. Moves are given by the throws of four
cowries with the usual values.
6.3.6. Somaliland (coastal regions): Tammam (Marin, a, 510). Played
on the same 5X5 board, but the middle row and column are rather wider
than the other rows and columns. The central cell is habarti, ‘mother’, and
the other crosscut cells are ‘ayn, Ar. ‘eyes’. The plain cells are dafandaf. Moves
are given by the throws of four cowries with the usual values, but the Arabic
terms for the numbers are used. A throw of 4 gives another throw, one of 8
gives two extra throws, but a second throw of 4 does not give a third throw
and a third throw of 8 in the same turn of play cancels all three throws. A
capture gives another throw, the capturing man is suldan, ‘sultan’, and the
captured man is ‘abid, ‘slave’ and has to be re-entered.
6.3.7. Southern India: Gavalata (Culin, d, 851). Played on a 5x5 board
I The same board is used in Celebes for the game gala (4.1.12) which Matthes describes as a
kind of backgammon although his description does not mention dice. It is possible that the
present game is also played in Celebes.
 <4) } a MerSa ({420}

132 RACE-GAMES [CH.6

(Fig. 58) with a larger number of crosscut cells. Two or four players, each
using one or two cowries as men. The moves
are given by four or five cowries but the values
are not stated. When two play, they enter their
men at A and B and proceed clockwise round
the board. Single men are taken and have to
re-enter, but the crosscut cells carry immunity
nu from capture. Culin implies that the board may
) « ¢ be any odd-sided square. FIG. 58. GAVALATA
Fertve 6.3.8. Japan: Do-chi-sugoroku, ‘travelling
gop * sugoroku’ (Prof. Tsuboi). Played on a 5x5 board without crosscut cells, See
‘the central cell being known as agori. agar 1—+tetihish

4. GAMES ON CRUCIFORM BOARDS

These games are also of Indian invention and were very popular under the
Mogul Dynasty when large boards were marked out in their palace court-
yards at Fatehpur-Sikrim, Agra, and Allahabad! on which the Mogul
emperors played with living persons as pieces. The board is now usually
embroidered on a piece of cloth. It consists of four arms (which I shall call
N., W., S., and E.), each containing 3 x 8 cells and arranged to form a cross
with a central enclosure known in Hindustani as char-koni, ‘the square’. On
each arm certain cells, chik, ‘fort’ or ‘castle’, are crosscut and men standing on
these cells are usually immune from capture. As a rule four persons play,
forming two sides, and sit at the outer ends of the arms, partners sitting
opposite one another. Each player has the same number of men which are
coloured to distinguish the sides, red and black opposing yellow and green.
Each player has his own track, starting from the central enclosure and passing
down the middle row of the arm against which he is seated, then returning
up the right-hand outer row of this arm and round the outer rows in an
anticlockwise direction to the end of the arm down which the track started,
and finally up the middle row to the enclosure again. When returning up the
middle row of this arm, the men are placed on their side to distinguish
them from men coming down the row. In some games the men are arranged
on different points of their track to shorten their journey.
In the older games the moves are given by the throws of two four-sided
dice (pasa), and in the later games by the throws of six or seven cowries. The
values of the throws vary in different games and places, but usually are as
follows:
Number of mouths up . NU Ome tenons Anes 8G
Value with seven cowries ~ (he AO 2 eae 4 oS” BOT ane
Value with six cowries . W 6.) 10-12 330 Anes) ae

Throws which are asterisked also give a grace, i.e. an extra 1 which may be
? A similar board at Delhi was destroyed after the Indian Mutiny.
6.3.8] RACE-GAMES 133
used separately to enter, re-enter, or bear a man. A player in his turn of play
continues to throw the cowries until they give a throw of 2, 3, or 4 when
he plays his men in agreement with his throws, and his turn of play ends
unless a capture gives him another turn of play. The opponent then has his

FIG. $9. CHAUPUR; PACHISI

turn of play. In most of the games a player may refuse to throw when it is
his turn to do so, or may throw and then refuse to use his throws. Throws
that cannot be used are lost.
Games on cruciform boards have spread from India to Ceylon, the Mal-
dive Islands, Burma, Sumatra, Persia, the Arabs of Palestine, Spain, and
Somaliland. They were introduced into Europe about 1896 on a smaller
board and named Judo, and K. C. Murray tells me that ludo has spread
recently from Europe to the Africans of Lagos, Nigeria.
6.4.1. India: Chaupur; Skr.: chatush-pada, ‘four square’ (Hyde, 68).
The crosscut cells are (S.) 5, 6, 12, 15, 69, 72, and corresponding cells on the
other arms; men on these cells are immune from capture. Four players
134 RACE-GAMES [CH. 6
forming two sides, each having four men, placed (S.) on 6, 7, 12, and 13
(Fig. 59) and on the corresponding cells on the other arms. Played with two
long dice, faces numbered 1 and 6, 3 and 4.
6.4.2. Ceylon: Pahada keliya, ‘race-game’ (Parker, 611). Cells on the
outer rows are called ge, ‘house’, and on the middle rows kamara, ‘room’.
Crosscut cells are (S.) 8, 9, 12, 72, 75, and corresponding cells of the other
arms, but carry no immunity from capture. Two, four, or eight players,
forming two sides; when four play each has four men (ittd, pl. ittd); if eight
play, each has two men. When four play, the men are placed on (S.) 6, 7,
and two on 12, and on corresponding cells of the other arms. Played with
two long dice, numbered 1 and 6, 3 and 4. Men can be doubled and are then
immune from capture. Single men can be taken and have to be re-entered,
but no special throw is necessary for re-entry.
When a player (say S.) has advanced all his men to 74, 75, and 8 (the three
cells immediately preceding his kamara cells) he has to wait there until his
partner (N.) has doubled two of his men on 68 and has then thrown doublets
and used this throw to advance these men to a cell behind 74. As soon as this
happens, S. can play and advance his men to his kamara cells. If, however, N.
had thrown doublets which he cannot use to advance from 68, or any other
throw, he must use these throws to advance his two backward men, but
these cannot advance beyond 67 and must wait until N. has thrown a doublet
which enables the two men on 68 to advance. Until this happens, the back-
ward men are liable to capture by E. or W. It is this special feature which in
Parker’s opinion entitles pahada keliya to rank as one of the best of all race-
games.
6.4.3. India: Chausar (Culin, d, 855; Falkener, 264). Crosscut cells are
(S.) 12, 72, and corresponding cells on the other arms, but these carry no
immunity from capture. Two or four players forming two sides, black and
yellow against red and green. All the black men must be borne before the
yellow and all the red before the green. When four play and one player has
borne all his men, he continues to throw but his partner uses the throws.
Played with three long dice, faces numbered 1 and 6, 2 and 5. When two of
the dice show the same number and the player has two men doubled, they
count as one man for this throw and are moved together by the doublet.
Falkener says that the rules of chausar vary from place to place in India.
6.4.4. Tamils, India: Tayam sdnalu (pronounced chénalu; Parker, 617).
Crosscut cells play no part in this game. Played by two persons, each with
eight men, or four, playing as partners, each with four men. Played with
two long dice, one numbered 1 and 4, 0 and 2, and the other 1 and 4, 0 and
3. Throws of 1, 0 (tayam), 4, 0, and 4, 4 (sd-nal) give a second throw.
The game starts with all the men in hand and no player can enter his first
man until he throws taéyam, which enters one man on the first cell of his
track. To enter each of his remaining men or a captured man, a throw of 1, 0
or 4, O or 4, 4 is necessary, any of these moves placing a man on its first cell.
6.4.4] RACE-GAMES 135
No man may pass over an enemy man in the course of play, and doubled
men on the same cell may be taken as if they were single men, but then both
are re-entered together.
6.4.5. India: Chauput (Falkener, 263). Chauput only differs from 6.4.4
in being played with cowries, but Falkener does not say how many. There
are, however, no throws of 10 or 25 and no graces. So long as a throw can be
used, either to enter or to move a man, it must be played. Two or four per-
sons play, forming two sides, and each player has four men. The initial posi-
tion of the men varies from place to place, and Falkener gives the following
alternatives: (S.) 6, 7, 12, and one in hand; 6, 7, 22, 24; two men on both 6
and 7; 6, 7, 72, and 74; and corresponding cells on the other arms.
6.4.6. India: Pachisi, ‘25’ (Herklots, App. viii; Falkener, 257; Culin, d,
851; Parker, 619). Crosscut cells are (S.) 8, 12, 72 or 8, 13, 71, and corre-
sponding cells on the other arms, and men on these cells are immune from
capture. Usually, four persons forming two sides play, each with four men
(goti, ‘horses’, in the Deccan; sari in Hindustani; kay in Tamil), but two per-
sons occasionally play, each with eight men. The moves are given by the
throw of six or seven cowries; according to Herklots, the values are those
given above, but Falkener (p. 261) gives for six cowries:!
Number of mouths up. 1 O. 4 Mae deseo
Value with six cowries . P26" VIOser gaged aoe

Throws of 10 and 25 give a grace, and a player continues throwing until he

throws a 2, 3, 4, or 5. A capture also gives an additional throw.
When the game begins, all the men are in hand. The first man is entered
by any throw, but the others can only be entered by a grace, and a grace is
necessary to re-enter a captured man. Men can be doubled on any cell but
doubled men can be taken by an equal or larger number of men. When
bearing, the exact throw to bring the man into the enclosure must be thrown.
A player can refuse to throw in his turn, or can throw and then refuse to
play the throw.
6.4.7. Tamils, India: Sokkattan; Ceylon: Pachis (Parker, 619). Crosscut
cells (chira) are (S.) 2, 4, 12, 72, and corresponding cells on the other arms;
they are cells of safety. The other cells are called ghora in Hindustani and vidu
in Tamil. Four persons, forming two sides, play, each having three men, or
three persons each with four men, or two each with six men. Played with
six cowries with the standard values. Throws of 10 and 25 carry a grace and
captures give another throw; if a player throws 10 or 25 three times running,
all these throws are cancelled but the player continues to throw. The game
begins with all men in hand and a grace throw (which places a man on the
first cell) is necessary to enter a man or a capture.
6.4.8. Burma: Pasit, chwe pyit-thi, ansah pyit-thi (Shway Yoe, ii. 83;
! Falkener also gives values for seven cowries, but as he arranges them in order of magnitude
and not by the number of mouths up, they are not worth repeating.
136 RACE-GAMES [CH. 6
MVB.). Crosscut cells are (S.) 12, 72, and corresponding cells on the other
arms. They are coloured red or green and called poh or kyah; they are cells
of safety. Four persons usually play, each with four men. Played with six
cowries (chwe), throws having the standard values. Throws of 10 and 25
carry a grace, and of 10, 12, or 25 give another throw. The cowries are
thrown into a shallow shell and if a cowrie falls outside the shell, that and
the preceding throw are cancelled. If three 10’s or three 25’s are thrown in
succession, all three throws are cancelled. Each player has three throws for
his first turn of play, but only one throw (with any extra throws gained) in
later turns of play.! When all the men are in play and one player has borne
all his men except one which is moving up its middle row to be borne, his
opponent can take it by playing a man to an adjacent cell on either outer
row of this arm.
6.4.9. Achehn, Sumatra: Pachih (Snouck Hurgronje, ii. 201). Played by
men only. Crosscut cells are (S.) 2, 11, 73, and corresponding cells on the
other arms. It is only on these cells that men (pawoih) can be doubled, but
they obtain no immunity from capture. Two, three, or four persons play,
each with four men, and there are no partners. Played with seven cowries
with the following values:
Number of mouths up > Oo cl: 2a tates 6 7
Valuesof the throws . - a? 10. Beaty A wos On TAS

The four throws asterisked give another throw, but a player cannot throw
more than three times in a turn of play, and his turn ends if he throws 2, 3,
or 4. The men are all placed on the first cell of their tracks and captured men
are re-entered on this cell, but no special throw is needed for re-entry. After
a player has made all the throws permitted in a turn of play, he adds up the
throws and moves one man that number of throws. Apparently there are no
grace throws in this game.
6.4.10. Maldive Islands: Dhola (Culin, d, 856). Crosscut cells are (S.) 4,
8, 13, 71, and corresponding cells on the other arms. Four persons play, each
with four men. Played with five cowries, but the values are not given.
6.4.11. Persia: Pachis, pichas (Culin, d, 856, but no particulars are given).
6.4.12. Arabs of Palestine: Barjis (Hilmi Samara); Spain: Parchis.
Played by women and children only. Crosscut cells are (S.) 11, 73, and
corresponding cells on the other arms; they are cells of safety. Four persons
play, each with four men. Played with six cowries, the throws having the
standard values. Throws of 10 and 25 carry a grace (khal), and men can only
enter and re-enter by a grace. Men can be doubled on any cell and doubled
men cannot be taken. If a player bears all her men before the opponents
have borne one man, the game is won by mars, counting seven games.
6.4.13. Somaliland: Bakkis (R. F. Burton, First Footprints in East Africa,
1856). The Indian pachisi.
? The game may also be played with three long dice as in 6.4.3.
6.4.14] RACE-GAMES 137

6.4.14. Madras (Hindus): Dajam (MVB., i. C. 5108b). Played on the

board of Fig. 60 with arms of 3 x § cells, the two end cells of the middle row

Pee ee bis Lal

Dee pedis xcS
cai dea | a alee || aa&

xX
FIG. 60. DAJAM

of each arm being crosscut. Played with cowries, but the number is not
given; the number of mouths up giving the value of the throw and no
mouths up counting 12.

FIG. 61. ASI KELIYA

6.4.15. Ceylon: Asi keliya, ‘the shell game’, or sonaru, ‘four-four’

(Parker, 614). Four persons play, the opposite pairs being partners, each
having four men, on the board of Fig. 61 with arms of 3x4 cells. Played
138 RACE-GAMES [CH. 6
with six cowries of which five (bella) are the common yellow cowries and
the sixth (so-bella) is a white cowry. The values of the throws are given by
the number which fall with mouth upwards. A throw of four mouths
upward is known as so-hatara when the white cowry falls with its mouth
down. A throw of 1, 5, or so-hatara, is called dinuma, ‘win’, and gives another
throw. Otherwise, there is only one throw in each turn of play. The value
of a throw must be used as a whole to move a man, except in the case of
so-hatara which may be sub-divided into 1, 1, I, 1. To enter or re-enter a
man I or so-hatara must be thrown but after paying 1 for entry or re-entry,
the player may leave the man un-entered to a later move when any throw
may be used to move this man. When a man has reached the middle row and
is moving up it to be borne, it can only be moved one cell at a time by a
throw of 1 or so-hatara. So-hatara is also the only throw which allows a man
to jump over an enemy man; this is due to its alternative value of 1, I, 1, I
which enables it to take the enemy man and then move on beyond its cell.
Doubled men have no immunity from capture; a single man moved to the
cell on which they stand takes both, but in compensation men which have
been taken together may be re-entered together as though they were one
man.
6.4.16. Europe: Ludo. A modification of 6.4.6. made and patented in
England (no. 14636) about 1896. The board is shown in Fig. 62; there
are no crosscut cells, and there are four squares between the arms, called
yards in which the men awaiting entry are placed. Two, three, or four per-
sons play, who sit at the ends of the arms, each having four men in his yard.
Moves are given by the throw of a single cubic die; a throw of 6 enters a
man on A and it has to run round the edge of the arms in an anticlockwise
direction to B and then up the middle cells of the arm from which it is borne.
A throw of 6 gives another throw. An enemy man is taken by a man which
is played to the cell on which it stands and has to return to its yard. Men can
be piled and then can be moved as though they were a single man, but then
they can be taken as a single man. The player (or side) which first bears all
his men wins.
6.4.17. Druses of Syria: Edris a jin, ‘the genii’s game’ (Culin, d, 857).
Played on a cruciform board, each arm containing 4x8 cells, and the
extremities of the arms are joined by slant rows of eight cells. Four players,
each with three men, one of which is called chief and the others soldiers,
though all move in the same way. Moves are given by the throws of four
cowries.
6.4.18. Siam: Len doat (Capt. Low, 379). Played on a lacquered board
(see Fig. 63). The game clearly belongs to the chaupur type, and has features
in common with 6.4.14, but the arms are arranged to form a rectangle.
Capt. Low gives the following description:
Two persons take each a side of the board. There are five cowrie shells for a dice,
each of the players has three pieces—each throws in turn, and if no. 1 or no. § casts up,
6.4.18] RACE-GAMES 139
the thrower continues to throw and to play until another number turns up. The chief
object is to pass through all one’s squares to those of the opposite party without

FIG. 63. LEN DOAT

interruption, taking his men if they can be overtaken by throwing a corresponding

number and leaping over them if the number cast exceeds. He whose pieces are thus
first returned to the place whence they set out, wins the game. Nos. 6, 15, 17, 26 on
each side are castles, and the piece holding one of these cannot be taken. The pieces
140 RACE-GAMES [CH. 6
which have been taken are re-entered again by casting the dice. No. 1 enters one,
no. 5, the whole are out, but the intermediate numbers do not enter any.

5. SINGLE-TRACK GAMES
Boards for single-track race-games, set out on the floors of buildings or on
pavements, may be seen in ruined towns in Egypt, southern Europe, Syria,
Persia, India, and China, and G. Marin (b, 64) has recorded many examples

|
DT
TT

~,D

<< =
ey, YY B
Q
Q LY
o
Pp

F G H
FIG. 64. SINGLE-TRACK BOARDS (after Marin)
Tamil games: from Mahabalipuram A-D
Kanarese games: from Vijanagar A, F; from Badami A, E; from Cravanbelgula F.
Singhalese games: G, H.

from the ruined cities of Dravidian India, both Tamil and Kanarese. The
track for these boards is set out in a series of small holes or cells with several
turns at right angles which, in the cell tracks, are marked by crosscut cells
and each straight length of track between these crosscut cells contains four
cells.! We may infer from other Asiatic race-games that the form of lots used
in these games gave five different throws, and that the crosscut cells were
posts of safety on which men were immune from capture. Similar diagrams
are still used in southern India and Ceylon.
6.5.1. Southern India and Ceylon: Pancha, panche, ‘the five game’
(Ludovisi, 35; Parker, 609). The board has the same shape as Fig. 64c, but
the number of cells in a and b is four, not five, and in c is six and in the
The terminal lengths may contain five or six cells.
6.5.1] RACE-GAMES I4I
preceding slant length is five. The circles outside the track are used for men
awaiting entry. Crosscut cells are called ge, ‘house’, and plain cells are kamara
or kattiya, ‘room’. Played by two, four, six, or eight persons, forming two
sides. Each side has three men (ittd, pl. itts) and six cowries are used as lots,
the value of a throw being the number of mouths upward. A throw of I, 5,
or 6 is necessary to enter, re-enter, or bear a man, and the player continues
to throw so long as he throws 1, 5, or 6, but throws of 2, 3, or 4 end the
turn of play. One side enters at a, the other at b, and both bear from c. As
in all these games, the exact number must be thrown to bear men.

Pi AGVis
[ ~' }
YUYY
c
J

FIG. 65. KAWADA KELIYA

6.5.2. Ceylon: Kawada keliya (Culin, d, 851). Played by two or four

players forming two sides on the board of Fig. 65. Each side has three men
(balo, pl. bala, ‘dog’). Six cowries are used with the same values of throw as
6.5.1. and the game is played in the same way.
6.5.3. Ceylon: name unrecorded (Marin, b, 64). The board is Fig. 64H.
142 RACE-GAMES [CH. 6
6.5.4. Korea: Nyout, nyout-nol-ki (Culin, a, 66; d. 681). Two persons
play with one or three men each, or three with three men, or four with four
men, forming two sides. The board (Fig. 66) is a circle with two diameters
at right angles. The men are called ma, ‘horses’. They are entered in the holes
A
Om ae)
O O
O O O
O O O

BOrsO,
“OVTO- 40, tome On
O O O
O z O
O O
On ee
C
FIG. 66. NYOUT

on the arc A, B, and move round the circle anticlockwise and a man reaching
B, C, or D can return to A by the diagonals, and are borne from A. The
moves are given by the throw of four staves, one side flat and white, the
other curved and blackened; their values are given by the number of white
sides upward and no white upward counts 5. Throws of 4 or § give another
throw, and all the throws are made before any is played. Men may be
doubled on any hole and then move as a single piece. Men can be taken on
any hole and then have to be re-entered. The game is played for a stake.
Culin says that nyout is the favourite Korean game. He claimed that it is
the ancestor of all board-games, even including chess, but this isabsurd. We do
not know whether it is played outside Korea or how old it is. Presumably
Korea obtained it from China, but no observer has seen it played in China.
6.5.5. Italy: Giuoco dell’ Oca; Spain: Juego de la Oca; France: Jeu de
l’Oie; England: Game of Goose; Holland: Ganzenspel ;Germany: Ganse-
spiel; Denmark: Gaasespil (Carrera, II Gioco de gli Scacchi, 1617, p. 25:
R. Holme, a, 68; Encycl. Méthod.; Strutt, 437; Culin, d, 841). Invented in
Florence under Francesco dei Medici (1574-87) who sent it to Philip II of
Spain, and it rapidly spread to other parts of Europe. It reached England in
1597, when John Wolfe “entered . . . the newe and most pleasant game of the
Goose’ in the Stationers’ Register, 16 June 1597 (Arber, iii. 21). In 1758 the
Duchess of Norfolk planted a game of goose in hornbeam at Worksop
(Horace Walpole, Letters (1840, iii. 395)).
6.5.5] RACE-GAMES 143
The game is played on a spiral course of sixty-three points, consecutively
numbered, which is set out on an oblong board. Any number of persons
can play, each with a single man which is entered on 1 and borne from 63,
the moves being given by the throw of two dice. A man can only be borne
when it comes to rest on 63 and if a higher number is thrown than that
which is necessary to reach 63, the surplus is played backwards from 63;
thus, with a man on $7 and a throw of 6, 3, the man is moved forwards to
63 and then back to 60.
Points 5, 9, 14, 18, 23, 27, 32, 36, 41, 45, 50, $4, 59, and 63 bear a goose
(whence the name of the game). Any throw that brings a man to a goose is
repeated until it brings the man to a cell without a goose. The only excep-
tions are 63 when the man is borne, or when the player’s first throw is 4, 5
or 3, 6 which would bring the man at once to 63. In these cases, a throw of
4, 5 enters the man on 53 and a throw of 3, 6 enters it on 26, and these points
bear two dice, showing 4, 5 and 3, 6 respectively.
Other points bear other signs and men reaching these points pay one to
the pool and, with one exception, incur other penalties: 6, a bridge, the man
is advanced to 12; 19, an inn, lose two moves; 31, a well, wait until another
man reaches this point and the two change places;! 42, a maze, go back to
30; §2, a prison, wait until another man reaches this point and the two change
places; 58, a skull, go back and start again.
There are no captures in this game; when a man reaches a point which is
already occupied by another man, the two men simply change places.
6.5.6. England: Snake (R. Holme, a, 68; Strutt, 437). The track is
shaped like a snake and composed of sixty-three holes for pegged men, the
holes being numbered 1-63. The moves are given by the throws of two dice
and men are entered on 1 and borne from 63. When a man reaches a point
already occupied by another man, it takes this man and it has to re-enter.
The game is probably now obsolete.
6.5.7. England: Snakes and ladders. Played on the 8x8 board, the
track starting at ar and running left to right along the odd rows and right to
left along the even rows, ending at h8 from which the men are borne.
Certain cells are joined by ladders and snakes. A man reaching the foot of a
ladder is placed on its summit, and a man reaching the head of a snake is
returned to its tail. The moves are given by the throws of two dice. A
favourite game with children.
6.5.8. Sudan: Li‘b el-merafib, ‘the hyaena game’ (R. Davies, SNR., viii.
145). Played by children on a board of 12x 12 cells. The track starts from
one corner, the village, and runs spirally through all the cells (days) to the
middle of the board (the well). Any number of children may play, each
having one man (his mother) which goes to the well to get water and returns
to the village. Moves are given by the throws of three staves, one face curved
and green, the other flat and white. The values of the throws are given by
T Strutt places the bridge on 5 and the well on 30, and adds 61, a goblet, add one to the pool.
144 RACE-GAMES [CH. 6
the number of white faces upward, and no white up counts 6. One white
face up is known as taba with no assigned value, but a throw of taba is neces-
sary at certain stages of the game and each player keeps a note of the number
of tabas he has in hand. A player goes on throwing until he throws 2, after
playing which the turn of play ends. To leave the village, a throw of taba is
necessary, and the exact throw must be obtained to reach the well, but tabas
(counting as 1) may be used to make up any odd number of days short of
the well. At the well, two tabas are used to drink, one to wash clothes, and
two to start on the return journey. If the player has insufficient tabas in hand,
he must wait until he throws them, but all the throws are noted and can be
used on the return journey. The first player to reach the village then releases
the hyaena, which requires two tabas to leave the village, but henceforward
it doubles the value of the throw. At the well it requires ten tabas to drink
and start back. On this return journey alone, it can eat any mother that it
overtakes.
6.5.9. China: Tsun k’i (Hyde, 65 has coan ki; Culin, b, 498). Played on
a square board divided into eight rows. Two persons play, each with sixteen
men, one having eight men on the eighth and eight on the seventh row on
the left-hand side of the rows; the other with eight men at the right-hand
edge of the first and second rows. The men move clockwise round the board
and the first to return all his men to their original positions wins. The moves
are given by the throws of two dice and a throw of 1, 1 loses a man. If
doublets are thrown, one man is moved half the throw, or two the whole
throw.

6. PROMOTION GAMES

I include in this section a few games played in China, Korea, Japan, and
Tibet, some of which seem to have been constructed to explain the complex
organization of the state services in these countries, showing the different
grades and ranks of officials and the ways in which these officials may receive
promotion; and others give instruction in the Buddhist cosmology, showing
the various conditions of existence and the advance through these to Nir-
vana. All are played on latticed boards (though the lattice is often irregular);
some cells are cells of entry which determine the branch of service, and
others bear the names of the posts which are gained by a man on reaching
the cell, and instructions for the following move. No description gives a
clear picture of any game, and Hyde, who devotes ten pages to 6.6.1, admits
that he has left many obscurities. We only know the names of most of the
games.
6.6.1. China: Shing kun t’o, ‘promotion of mandarins’ (Hyde, 70, who
describes the game as played under the Ming dynasty, 1368-1616). Played
by four or more persons whose careers are to be illustrated by the game.
Each has one man whose moves give its owner’s career. The board contains
6.6.1] RACE-GAMES 145

ninety-eight cells and the track runs round the board three times and then
up to the central square on which stands the imperial palace; twenty-one

ae)N

Pees =
hon
nho> N 76 78 |79 32

|
8

ri
\A 33

le tee

||" eee
Ps s
|e 36
OoO|™

o
ine)
—~}|—~ o
IN

FIG. 67. CHINESE PROMOTION GAME

cells are cells of entry, and seventy-seven form the track, each cell bearing
the name of the post filled by the man occupying it, and instructions as to
the next move.
The moves are given by the throw of six dice which are thrown into a
bowl that is placed in the middle of the board. The value of a throw depends
5356 r
146 RACE-GAMES ICEL
not so much on the actual numbers shown by the dice as on their distribu-
tion and the number of dice with the same number of pips. The highest
throw is six dice all showing the same number or six dice all different; only
when it is necessary to discriminate between two throws of six alike does the
nature of the throw (all 6’s, all 5’s, &c.) come in. Next in value is five alike,
then four alike, three alike, and two alike. Throws of two alike enter men on
cells 1-6, three alike enter men on 9-14, four alike on 16-21, five alike on 7,
six alike on 8, and another throw (which Hyde omits) on 15. After a player
has entered his man, the only throws that allow a move are four 4’s and for
a few cells, 3, 4, 5, 6, the actual move being that stated on the cell.
6.6.2. China: Shing kun t’o (Culin, b, 504). The game under the Manchu
dynasty which supplanted 6.6.1. The board contains seventy-six cells of
which thirteen are cells of entry. As in 6.6.1 each cell bears the name of a
post in the state services and instructions as to the next move.
The moves are given by the throw of four dice, and the value of a throw
depends, as in 6.6.1, on the distribution of the numbers on the dice, but, as in
other modern Chinese dice-games, the throw of 4 counts above the throw of
6. Men are entered by the values, 3, 4, 5, 6, then three 4’s, three 6’s, three $’s,
three 3’s, three 2’s, three 1’s, and then two alike in the same order down to
two 1's. Subsequent promotion depends on the throws, two alike move
once, three alike twice, and four alike, three times,! the actual method of
moving being given on the cell from which the move begins. Throws of
1’s (chong, ‘stupidity’) generally mean a move of demotion. Careful analysis
of the moves on Hyde’s diagram of the board shows that throws of 1 and 2
result in the degradation of the man to a lower rank.
6.6.3. Korea: Tjyong-kyeng-to, ‘game of dignitaries’ (Culin, a; d, 820).
The Korean variety of 6.6.2. The board is a lattice of 9 x 12 cells, each bearing
the title of a Korean official and directions as to the next move. The moves
are given by a long five-sided wooden die, the edges of which are notched
with from one to five nicks.
6.6.4. Japan: Kuwanroku (Culin, b, 504 n.). A similar but simpler game
with the titles of Japanese officials.
6.6.5. Japan: Tobi-sugoroku, ‘jumping sugoroku’ (Prof. Tsuboi).
6.6.6. Korea: Monk’s tjyong-kyeng-to (Culin, d, 821). A board of
13 X13 cells, containing the names of the various conditions of existence of
the Buddhist cosmology. It is a Buddhist promotion game, played like the
preceding games with three small wooden dice, the faces bearing the magic
formula nam mo o ni t’0 fat (Namah Amitabha) instead of pips. The cells, as in
the other boards, contain instructions as to the actual moves to be made.
6.6.7. Tibet: name unrecorded (Schlagintweit, Buddhism in Tibet, cited
in Culin, d, 821 with figure, which is repeated in HC. 43. The board is

? This may apply to 6.6.1 also, but Hyde only says that the player continues to throw until he
obtains a throw that he can use, but this seems fatuous. Possibly he meant that a player who
obtained a usable throw was given another throw as in other oriental games.
6.6.7] RACE-GAMES 147

one of 8x8 cells, but only the alternate rows are cells, the intervening rows
bearing text which probably gives instructions as to the move from the cell
above.
6.6.8. Tibet: Srid-pai khorio, “Circle of life’ (Sarat Chandra Das, 338).
In this game the players obtain different positions in the various heavens and
hells.

7. AMERICAN INDIAN RACE-GAMES

The use of staves and other objects as lots for gambling as well as for race-
games was already common in America before the Spanish Conquest (E. B.
Tylor, “On the Game of Patolli’ in JAI., 1878, viii. 116 and 1878-9, ix. 23,
and ‘On American Lot Games’ in IAFE., 1896; Culin, d, 800-2). The older
games are:
6.7.1. Ancient Mexico: Patolli' (Culin, d, 844 and pl. 29). Played by two
persons on the cruciform board of Fig. 68. The
board consists of four arms, each of 2x 7 cells,
with four cells in the centre where the arms
cross. On two opposite arms the bounding
lines between the second and third cells from
their outer ends, and on the other arms the
bounding lines between the third and fourth
cells from their outer ends are thickened to
form two triangles with their apexes pointing Frc. 68. PATOLLI (Duran)
inwards. Duran, Historias de las Indias, 1880, m1.
xxii, from whom the diagram is taken, says that each player had ten men,
and the diagram shows eleven men on cells, three lying without the board,
and one about to be entered by one of the players.
6.7.2. Ancient Mexico: name unrecorded (Culin, d, 802, pl. 17; from the
Fejervary Codex, Third Annual Report, Bureau of Ethnology, Washington).
The painting, of which Fig. 69 shows the scheme, is generally assumed to
show a game-board. The twenty points where the track changes direction
are marked by circles, those at the ends of the four narrower arms being
larger than the others. On each stretch of the track between two circles are
twelve points (omitted in the figure), so that the track contains 260 points.
Culin (d, 803) calls the board ‘a divinatory, calendrical counting circuit’, and
I am very doubtful if it is a game-board at all. The centre square of the picture
contains the figures of a warrior-god or -priest, armed with atlat] and darts.
6.7.3. Aracaunian tribe, Chile: Quecho (Molina, Hist. Chile, m. x, cited “
in D. Forbes, History of Chess, London 1860, p. lvi). “Quecho has a great
affinity to backgammon, but instead of dice, they make use of triangular
™ Patolli, from the verb patoa, ‘to play a game’, seems to have been a generic term for any game.
Thus, when the Mexicans learnt card-games from the Spaniards, they called them amapatolli,
‘paper-game’, and chess was called quaah patolli, ‘wood-game’ (Culin, d, 935). Aztec dictionaries
also give cauallopan, ‘the cane-games’, and nemininaliztli, ‘piece (lit. horse) games’ (Culin, 4,
800-1).
148 RACE-GAMES (CH. 6
pieces of bone marked with points which they throw with a little hoop or
circle, supported by two pegs.’

FIG. 69. BOARD FROM THE FEJERVARY CODEX (Culin)

The existing race-games of the American Indian tribes in Central and

North America have been collected by S. Culin in his Chess and Playing
Cards, 1898, and Games ofthe North American Indians, 1907, and the following
descriptions of these games have been taken from these works.
6.7.4. Shoshonean stock, Hopi Indians, Oraibi, Arizona: name unrecorded
(Culin, g, 161). The board (see Fig. 70) is marked on a slab of sandstone. It
is oval in outline with small circles at either
end, then five transverse lines which mark the
\S boundaries of the ‘points’, and a small circle in
) the middle.
QO 6.7.5. Shoshonean stock, Paiuti Indians,
Pyramid Lake, Nevada: Tatsungin (Culin, g,
FIG. 70. STONE BOARD _167). The board is marked on the ground by
FROM ORAIBI (after Culin) a row of ten sticks, the points being the
intervals between the sticks. Two men play,
each having one piece which is entered at opposite ends of the line, and
moves along the line. The moves are given by the throws of eight staves
(split pieces of reeds with the flat face coloured red) and the value of the
throw is the number of staves which fall with the white face upwards.
6.7.6] RACE-GAMES 149
6.7.6. Mayan stock, Ketchi Indians, North Guatemala: Boolik, ‘dice
play’ (Culin, g, 141). The board is marked on the clay floor of houses by a
row of fifteen grains of corn, the points being the intervals between the
grains. The game is played by the light of the fire. Any even number of
persons play, forming two sides, and each player has five bits of leaf or stick
of different lengths to differentiate the pieces belonging to him. The players of
each side play in turn alternately, and the moves are given by the throws of
four grains of corn, one side of each being flattened and marked by a dot.
The value of the throw is given by the number of dots falling upward and
none up counts 5. Each player has two throws in a turn of play. The sides
enter at opposite ends of the track and on reaching the other end a man
returns to its starting-point (as though the track were continuous) and
traverses the track again. Each player can only have one movable man at a
time on the track, and when that man is captured, he enters another man.
Whenever a player reaches a point that is occupied by an opponent, he adds
his own man to it and moves them together in the opposite direction until
they reach his starting-point, when he captures the opponent and re-enters
his own man for a fresh journey along the track. During journeys in the
reverse direction, the combined men are vulnerable and if any of the oppo-
site side plays a man to the point occupied by the reversing men, he adds his
man to the pile and moves it back to his starting-point and on arriving there,
he takes the opponent man and liberates those which belong to his own side.
The game ends when all the men of one side are taken.
6.7.7. Mayan stock, Chichen Itza Indians, Yucatan: name unrecorded
(Culin, g, 143). A similar game to 6.7.6.
6.7.8. Mariposan stock, Yocuts Indians, Tule River reservation, Tulare
County, California: Tsikehi, ‘to hurdle’ (Culin, g, 140). The board is
marked on the ground by a row of twenty-five sticks, the points being the
intervals between the sticks. Each player has four horses (witchet) and the
moves are given by the throws of four staves
with one side flat and one round. One flat
up = I; two flat up = 2; three flat up = 3;
and four flat up = 16.
6.7.9. Shoshonean stock, Hopi Indians,
Oraibi, Arizona: Totolospi! (Culin, g, 160).
The board (see Fig. 71), marked on a slab of
sandstone, is cruciform with four arms, each pyc, 71, STONE BOARD
with a circle at its outer end on which men’ FROM oratst (after Culin)
are entered and five transverse lines which
separate the points of the tracks, and with a circle where the arms cross
for the terminals of the tracks. Two or four persons play, each having
one man and the moves are given by the throws of three staves with one
' Totolospi is also the name of the war-game, 4.1.15. Cf. the Aztec totoloque, which is used for
a dice game.
150 RACE-GAMES [CH. 6
face flat and unmarked and the other round with a pattern of dots. The
only throws that have a value are those in which all the staves fall in the
same way = I, and the player continues to throw until the staves fall in
different ways.
6.7.10. Shoshonean stock, Hopi Indians, Oraibi, Arizona: Totolospi
(Culin, g, 161-2). Three boards on sandstone slabs are figured (see Fig. 72);

o IlIl@IIII| ©

FIG. 72. STONE BOARDS FOR TOTOLOSP! (Culin)

A was collected by Culin and B and C by Rev. H. R. Voth; A and B were

used by two or four persons, C by a larger number, each player having one
man which is entered in the circles on the arm running up the middle and
traversing the square track anticlockwise according to Voth, or in the circles
on the square track according to Culin. Moves are given by the throws of
two or three staves, one side white and the other figured. The only throws
that have a value are those in which all the staves fall in the same way = 1
(Voth) or figured = 1 and white = 2 (Culin). The player continues to throw
until the staves fall in different ways. Voth noted that a man was sometimes
advanced more than one point or returned to its starting-point, but failed to
ascertain why.
Most of the American Indian tribes in U.S.A. and Mexico play, with
small differences in rule, a game which is said to have been originally divina-
tory, as it still is with some tribes. The board is sometimes marked on a
piece of cloth or hide, but more often set out on the ground in the form of
a hollow circle or square, from 3 to § feet across. In the centre of the diagram
is placed a flat stone, called the ahl stone, and in throwing the staves they
must hit this stone and bounce off it. When the board on the ground is
circular, the circumference is marked by forty stones at equal intervals along
it, except that each quadrant of ten stones is separated from its neighbours
by a larger interval which is known as a gate, door, river, or creek and
6.7.10] RACE-GAMES 151

marks a point of entry for the men. The board is usually orientated so that the
gates occupy the cardinal points. The points on which the men move are
the smaller intervals between the stones.
When the board on the ground is square, the outline is marked by small
holes in the ground in which the men are placed and move. The corner holes
are the gates and, excluding the gates, there are eight holes along each side
of the square. In some games the gates are approached by outer arms of five
holes.
Two persons, or four forming two sides, usually play, and each side has
one or two men, commonly known as horses, which are entered on the track
and traverse the track a prescribed number of times and finish at the gate by
which they entered. There is no invariable rule as to the direction in which
the track is traversed; both sides may move in the same direction or in
opposite directions, but the direction once chosen cannot be changed in the
course of a game. Some games are played without captures, but generally a
man has to return to its starting-point when an opponent is played to the
point on which it stands. A man occupying a gate acquires no immunity
from capture thereby; indeed in some games it is penalized by being sent
back to its starting-point.
The moves are given by the throw of three or four staves, and it is conve-
nient to classify the games by the number of staves used. The staves are
usually pieces of split cane, one side flat and the other round and often also
coloured, or the staves may be flat on both sides, the sides being coloured
differently. Occasionally one stave is notched when three are used. The
values of the throws may vary but the usual values are:
(a) Three staves with flat and round sides. All round up = 10 and another
throw; two round up = 3; one round up = 2; no round up = 5.
(6) Three staves, one notched. All round up = 10 and another throw; two
round up, one notched = 2; two round up, neither notched = 1; one round
up, not notched = 3; one round up, notched = 15 and another throw; no
round up = 5.
(c) Four staves, one face flat and the other round. No round up = Io and
another throw; one round up = 1; two round up = 2; three round up = 3;
four round up = 5 and another throw.
(a) THREE STAVES
6.7.11. Tanoan stock; Tigua Indians, Isleta, New Mexico: Patol (Culin,
d, 763; g, 190). Board a circle, forty stones, four rivers (p’ayhiah). One circuit.
Any number of players playing individually, each with one horse (kahniddeh).
Each player chooses his river for entry and direction of move. No penalty
for falling in a river. Captures as usual but children often play without cap-
tures. Also played with a notched stave (see 6.7.24).
6.7.12. Tanoan stock, Tigua Indians, Isleta, New Mexico: Cuwee (Culin,
g, 190). The same game as 6.7.11 but played by boys without captures.
152 RACE-GAMES [CH. 6
6.7.13. Keresan stock, Keres Indians, Laguna, New Mexico: Kawasukuts
(Culin, d, 729; g, 122). The same game as 6.7.11 but the rivers are called
siamma, ‘door’, and a player who falls into a river (other than his starting-
point) has to go back and re-enter.
6.7.14. Tanoan stock, Tewa Indians, Hano, Arizona: name unrecorded
(Culin, g, 190).
6.7.15. Athapascan stock; White Mountain Apache Indians, Arizona:
Satill; Tze-chis’ or Zsetilth’; Setichch’ or Tsaydithi (Culin, g, 87, 90).
Board a circle or square, forty stones, four gates. One circuit. Played by
women; two, three, four, or six players forming two sides, partners sit
opposite one another. Each side has two men. The sides move in opposite
directions. Gates are starting- and finishing-points only. Captures as usual.
6.7.16. Athapascan stock, White Mountain Apache Indians, White River,
Arizona: Sét dilth’ (Culin, g, 88). Board a circle with four gates, NE., SE.,
SW., NW., and all men enter by the NE. gate. Three circuits. Two, three,
or four players forming two sides; partners sit opposite one another. A man
which reaches a gate is placed on the point behind the gate. Captures as
usual, and a man is also taken if an opponent man gets in advance of it.
6.7.17. Athapascan stock, Navaho Indians, New Mexico: Sitih (Culin,
g, 95). Board a circle, forty stones. Played by women.
6.7.18. Athapascan stock, Navaho Indians, New Mexico: Settilth (Culin,
g, 94). Board a circle, forty stones. Throws of two or one round up are
without value. Played by women.
6.7.19. Athapascan stock, Navaho Indians: Tsidil, Tsindil (Culin, g, 94).
Board a square, each side ten stones but the corner stones are counted twice,
once for each side. Played by women.
6.7.20. Zufiian stock, Zufii Indians, Zufii, New Mexico: Tasholiwe,
‘wooden canes’ (Culin, d, 773; g, 221). The secular form ofSho’liwe (6.7.29).
The board a circle, forty stones and four gates, the board being so orientated
that the gates are at the cardinal points. Four circuits. Two or four persons
play each with one man, represented by a coloured splint. They are placed
on the gates, yellow N., blue W., red S., and white E. Yellow and blue move
from right to left and red and white from left to right. Captures as usual.
6.7.21. Keresan stock, Keres Indians, Acoma, New Mexico: Inaani
(Culin, g, 120). Board a circle, thirty stones and three doors (tsiama). One
circuit. Also played with a piece of bone, one side black, the other white
with values, black side up = 10 and throw again; white up = 5.
6.7.22. Yuman stock, Havasupai Indians, Arizona: Hue-ta-quee-che-ka
(Culin, d, 768; g, 200). Board a circle of forty stones and only one gate at
which the men enter. Any number of girls play, forming two sides, playing
in opposite directions from the single gate.
6.7.22a. Yuman stock, Walapai Indians, Arizona: Tawfa (Culin, g, 208).
Board a circle, but only about four-fifths of the circumference is used. Half-
way round the used part is a larger stone and the arcs on either side are
6.7.22a] RACE-GAMES 153
marked by twenty-five smaller stones. Played by women forming two sides.
Men entered in the wider gap and the two sides move in opposite directions
and the first side to bring its men to the larger stone wins. The game was
falling into disuse about 1900, being replaced by cards.

(b) THREE STAVES, ONE NOTCHED

6.7.23. Tanoan stock, Tewa Indians, Nambe, New Mexico: Tugea or
Patol (Culin, g, 192). Board a circle; forty stones and four gates at the
cardinal points. One circuit. Two or more persons play, each with one horse.
All enter at the same gate, and all move in the same or opposite directions as
each player chooses. Captures as usual, except on gates which may be occu-
pied by two or more horses.
6.7.24. Tanoan stock, Tigua Indians, Isletas New Mexico: Patol. Only
differs from 6.7.11 in having one stave notched.
6.7.25. Tanoan stock, Tewa Indians, Santa Clara, New Mexico: Tugi’-e-
pfe; in Mexico, Quince (Culin, g, 193). Board a square in sections of ten
grains of corn or pebbles. One circuit. Two persons play. All horses enter by
the same gate. Played in two ways: (1) without captures, all men moving in
opposite directions; (2) with captures, all men moving in the same direction.
Captures as usual, and, in addition, if a player by duplicating his first throw
can capture, he may do so, but his opponent when re-entering his men may
elect in which direction they move. If he chooses to move in the same direc-
tion, captures are possible, if in the opposite direction, captures cannot be
made.
6.7.26. Keresan stock, Keres Indians, Laguna, New Mexico: Owasokotz,
or Wasokuts (Culin, g, 121). Board a circle, forty stones and four doors
(siamma) at the cardinal points. Any number of players, playing individually
or as partners, each with one man, formerly opolaniama but now a horse.
All enter by the E. door and may choose in which direction to move. Played
in two ways: (1) pass without captures, but if a player’s first throw takes him
right round the track and back to the E. gate, he must start again; (2) enter
with captures.
6.7.27. Keresan stock, Keres Indians, Cochiti, New Mexico: name un-
recorded (Culin, g, 121). Board a circle; forty stones.
6.7.28. Tanoan stock, Tigua Indians, Taos, New Mexico: Caseheapana
(Culin, d, 765; g, 195). Board a circle of 160 stones, with larger ones dividing
the circumference into four quadrants, each of forty stones. The diameter E.
to W. isa river; that from N. to S. is a trail. One circuit. Two persons play,
each with one man. The men are entered from the river at E. and move in
opposite directions. Captures as usual, but only after the captured man has
passed the end of a trail.
The game is usually played on the night of ‘the day of the dead’, 3 Novem-
ber.
154 RACE-GAMES [CH. 6

(c) FOUR STAVES

6.7.29. Zuftian Indians, Zufii, New Mexico: Sho’liwe, ‘cane arrows’
(Culin, d, 782; g, 213). Formerly exclusively a game of war divination and
only played by the priests of the bow at the annual festivals of the Twain
gods of war, Ahaiynta and Matsailema. Board a circle, forty grains of corn,
four gates. Four circuits, made clockwise. Four persons play, each with one
horse. Captures as usual.
6.7.30. Kiowan stock, Kiowa Indians, Oklahoma: Zohn ahl, ‘creek

16 N 16
g 18 19 O
OO oO
20 20

15 15
140 O
130 O
120 O

W oS
i u

/0 10
2 0 O
ra) O
Ta) O

6 6

= §
299) /
9990/7
ic 5

FIG. 73. ZOHN AHL

wood’ (commonly called ‘the Awl game’ from the shape of the men,
Culin, d, 686; g, 124). Played on the board of Fig. 73. It is a square with the
corners cut off by arcs of circles (called ‘knees’) and is marked by forty
small stones, the points being the intervals between the stones. At the cardinal
points are wider gaps, N. and S. are rivers or creeks, and men falling in these
have to go back and start again; E. and W. are dry river-beds which involve
no penalty. Four circuits. Two or four players, each with one man for
6.7.30] RACE-GAMES 155

moving and four counters for scoring. When four play, they form two sides.
Men are entered at S. and the sides move in opposite directions. Captures as
usual. A player pays one counter for falling in a river or for re-entering a
man, and gains a counter for each completion of a circuit. If one side gains
all the counters before completing the four circuits, it wins outright. Only
played by women.
The Comanche and Arapaho Indians play a similar game with eight ahl
sticks (staves).
6.7.31. Kiowan stock, Kiowa Indians, Oklahoma: Ne’baku’thana (Ara-
paho); Tsona, ‘awl game’ (Kiowa) (Culin, g, 126). Played on the board of
Fig. 73 by women only, and only differs from 6.7.30 in the treatment of the
throws. One stave has a green line along its flat side and is known as sahe,
‘green’. If this stave falls with the flat side up, it gives another throw.
6.7.32. Tanoan stock, Tewa Indians, Santa Clara, New Mexico: name
unrecorded (Culin, g, 193). Board a square; forty stones and four gates at
the corners. One circuit. Each player has one man.
6.7.33. Keresan stock, Keres Indians, Sia, New Mexico: Wash’kasi
(Culin, d, 730; g, 123). Board a square, thirty-six stones, one being in each
corner. Two players, each with one man which is entered at the same corner,
and the players move round the track in opposite directions. The first of four
games played by Po’shaiyanne, the Sia cultural hero, against the tribal priest.
6.7.34. Yuman stock, Mission Indians, Mesa Grande, California: Serup
(Culin, g, 204). A wooden board marked with holes in the shape of a square
with the N. side extended E. with four extra holes and the S. side extended
W. in the same way. The square is made up of thirty-six holes (each side
has ten holes but the corner holes are counted twice over to make up the
ten holes in the two sides meeting at the corners).
6.7.35. Piman stock, Tepehuan Indians, Chihuahua, Mexico: Quince,!
Intuvigai zuli gairagai, ‘game straight throwing’; Zuaque Indians, Sinaloa,
Mexico: Kezute; Tarahumara Indians, Pueblo of Carichic, Chihuahua,
Mexico: Romavoa (Culin, g, 152, 153). Board a square of thirty-six holes
on a piece of wood or marked in the ground, with arcs of five holes branch-
ing from two opposite corners by which the men enter the square track (see
Fig. 74). One circuit. Two persons play, each having one man. The players
move round the track in opposite directions. Captures as usual.
6.7.36. Piman stock, Papago Indians, Pima County, Arizona: Chings-
koot; Pima Indians: Kinsgoot, Kints kut (Culin, g, 146, 150, 151). Board
as shown in Fig. 74. The semi-circular arcs are called kee, ‘houses’, and the
corner holes are utpa, ‘doors’. In the Papago game, two, three, or four per-
sons play as individuals or as forming two sides, and each player has two
horses which have to enter in turn by the arcs, and the second cannot enter
until the first has left the arc. When both horses are in play, either can be
! The Spanish word for ‘fifteen’, which has been adopted by the Indian tribes as the name
(kints) of the stave with the value 15.
 156 RACE GAMES [CH.6

moved. Each player can elect in which direction he goes round the track.
If A enters by a throw of 15, B, throwing 15 or 14 (for there is no throw of
1), can take A by entering a horse to move in the opposite direction. The
exact throw to move off the track must be thrown, but if the horse stands
on the last point, a throw of 2 takes it off. In the Pima game, each player
has only one horse.

6.010 00 "0 01001005

fe) O O
fe) O S
O Oo \
O O fe)
W E
O fe O
\o O O
is O O
> O fe)
ONO OO ONO NeOLOWOTONC
S
FIG. 74. QUINCE

6.7.37. Piman stock, Pima Indians, Arizona: name unrecorded (Culin,

g, 149). Board that of Fig. 74. Any number of persons forming two sides;
one man to each side. Men are entered on the middle hole of the arc and
play round the square and back to the point on which they entered. Players
move in opposite directions and captures as usual.

8. THE ASTRONOMICAL GAME

6.8.1. Arabs, Persia, Spain: Al-falakiya (al-Ma'sidi); Kawakib (al-
Amuli); Los escaques, ‘chess’ (Alf.)—(AIf. f.94a; HC. 343, 349 with figure
of board). Now obsolete. (murray Hist: A © hess
Played by seven persons, each with one of the seven luminaries of the
Ptolemaic system (1. Moon; 2. Mercury; 3. Venus; 4. Sun; 5. Mars; 6. Jupi-
ter; 7. Saturn). The board is a circular figure of seven concentric rings, each
divided into twelve houses by radii from the common centre. Each house is
allotted to one of the signs of the Zodiac, arranged clockwise round the
board. Each ring is occupied by one luminary in the order given above from
the centre outwards. Each house in each ring is divided into as many points
as it is distant from the centre. The moves are given by the throws of a
seven-sided die, a throw of 1 moving the Moon, 2 Mercury, &c. If'a luminary
6.8.1] RACE GAMES 157
moves in sextile to another luminary (i.e. is two houses distant) the
player wins 24 from the player of the other luminary; if in trine (i.e. three
houses distant) he wins 36; if in quadrature (i.e. four houses distant) he pays

itigc
ee

Ms,

1
RK Oo,

(i.e. six houses distan

he pays 72, and if in conjunction (i.e. the same house) he pays 12. The game
continues as long as the players choose to play, as there is no situation the
gain of which determines the result of the game."
For a Nigerian game, ashere, played on a similar figure of seven concentric rings, see note on
14.
 7
MANCALA PART I
Mancata (Arabic mangala or mingala, with the accent on the first syllable in
Syria and on the second in Egypt, from the verb nagala, ‘to move’) is the
name of a popular game in the Levant and Egypt which is played on a board
containing two rows of cup-shaped depressions or holes in which the pieces
are arranged and moved. Anthropologists use the term mancala for any simi-
lar game played on a board in which the pattern of lines and cells usual for
board-games is replaced by two, three, or four rows of holes deep enough to
contain a number of pieces at the same time. I shall do the same, but shall call
the games mancala II, III, and IV, classifying the games by the number of
rows of holes in the boards. In the present chapter I shall first deal with the
general characteristics of the games, their materials, and the evidence for
their antiquity, and next with the mancala II games, reserving the mancala
III and IV games for Chapter 8.
Today mancala games are widely spread over the tropical and sub-tropical
regions of Asia, Africa, and the adjacent islands, and African slaves carried
them to the West Indies and America. Mancala II is played in Asia, in Arabia,
Palestine (Arabs only), Syria, Turkey, Iraq, Iran, India, Ceylon, southern
China, Annam, Siam, Malaya, Indonesia, and the Philippines. In Africa it is
played in Egypt, Tunisia, Nubia, and the Sudan, in the Bahr el-Ghazal,
Abyssinia, Somaliland, and Kenya, in West Africa from the Senegal to the
Gabon, in the Congo basin and Angola. In America it is played by peoples
of African ancestry in all parts to which the slave trade extended. Mancala
III, now obsolete in Arabia, Abyssinia, and possibly in Somaliland, is still
played in parts of the Eastern Sudan and Kenya. Mancala IV is played in the
Sudan, Kenya, Uganda, Ruanda, Tanganyika, Zanzibar, Mozambique,
Rhodesia, Nyasaland, the Transvaal, South Africa, and Madagascar. The
popularity of mancala in Africa is so great and so general as to justify Culin
in calling it ‘the national game of Africa’. Although J. Galt (p. 242) saw it
played in the Cyclades in 1810, mancala has never taken root in the non-
Islamic parts of Europe.
The wide distribution today of mancala games and the varieties of ways
in which they are played make it difficult to trace the connexions of the
existing games, but their diffusion in Asia seems to have been from west to
east, and in Africa from north-east to west and south, both pointing to an
earlier practice of mancala in Egypt or Arabia, but we cannot say much
more than that. Sir H. H. Johnston (‘Survey of the Ethnography of Africa’,
in JRAI., xliii. 1913) has argued for a very early penetration from Egypt to
 MANCALA. PART I 159
Negro Africa of ‘domesticated animals, all musical instruments superior to
the musical bow and the drum, several types of games played with hollowed
or divided boards and counters, and a good many Egyptian notions about
religion’, and in support of this view is the fact that mancala II boards existed
in Egypt in the Empire Age (c. 1580-1150 B.c.), but we have no means of
filling the gap between this and the next appearance of the mancala II board
in Ceylon during the early centuries a.D. and in Arabia before the time of
Muhammad. On the other hand, there are indications that the mancala III
and IV games spread from southern Arabia by the Straits of Bab-el-Mandeb
or the southern end of the Red Sea to the opposite African shore, to Abyssinia
and Somaliland and thence inland to Kenya and Uganda, and southwards to
Zanzibar and Southern Africa.! In later times, Muhammadanism has carried
mancala II to most countries to which its religion and culture have extended.
In Syria, Egypt, and West Africa generally, mancala II is played by men,
women, and children, men usually only playing with men, and women with
women. Among the Tamils of south India, in Ceylon, Malaya, Indonesia,
and the Philippines, mancala II is predominantly a women’s game and only
played by men in less civilized or more remote regions. In central, east, and
southern Africa, mancala IV is almost entirely a man’s game. As Driberg
(p. 114) says: “women very rarely play and then only among themselves,
and they play it not with much skill or enjoyment, but apparently in more
trivial moods, inconsequentially, in the intervals of gossip.’
Most observers of the mancala games of the Old World regard them as
primarily a form of recreation which can be played after the day’s work is
finished, during the midday rest, or when rain prevents work. There seems
in Africa to be no season of the year when the game is specially appropriate,
but in Ceylon it is most played at the season of the New Year. Many people
will not play mancala after sunset, but this may be true of other board-games.
Some of the closest observers of mancala in the Old World assert that the
game has no social or religious associations. On the other hand, Herskovits
(b, 34) has found a semi-religious significance attached to mancala in the
West Indies and, most markedly, in the Surinam Bush, Dutch Guiana:
‘here it is definitely associated with death; it is the game which is played
in the House of Mourning to amuse the spirit whose body is awaiting
burial. The play is carried on during the day and this is why the game
is one to be played in the daylight hours only.’ He suggests that these
beliefs were brought by slaves from West Africa and that they point
to an earlier attitude towards mancala which is now obsolete or forgotten
in Africa.
Mancala is often played on boards which are marked out on the ground.
Doughty (Travels in Arabia Deserta, 1888, ii. 536) describes these at Teyma
™ Mancala III was already played in Arabia in the early days of the caliphate, and the simpler
forms of mancala IV in Zanzibar are called kiarabu, ‘Arabian’. There is no evidence that mancala
III or IV was ever played in Egypt.
160 MANCALA. PART I [CHF
in central Arabia, and Sanderson (p. 726) says that the characteristic holes in
the ground are to be seen in every Bantu village in Nyasaland. Sometimes
these open-air boards take a more permanent form; Partridge (p. 259) saw
at Awakanda on the Cross River, Nigeria, boards cut out on the projecting-
roots of trees on the beach where young men and boys forgathered to play
azigo. In Abyssinia (Cohen, 491) and Angola (Culin, c, 602) mancala boards
are often cut on outcrops of rock by the roadside. Some of these boards may
be old, though the surroundings in which they occur rarely give any indica-
tion of how old. Many such boards have been reported from Kenya and
Uganda, particularly round the Victoria Nyanza and farther south in
Rhodesia and the lower Zambezi valley. I may instance:
1. Some boards of 2x 13 holes in Kenya which Dr. Leakey assigned to neolithic
times.
2. Some boards near Ussindja and on Ukarere Island in the Victoria Nyanza,
described by v. Luschan (a, 881) and Kollmann (a, 77, 96; b, 108, 135).
3. A group of game-boards at the Muzizi Bridge-crossing on the Buganda-Toro
border, cut in sandstone and much weathered before they were covered by silt
which was again disappearing when Wayland (b, 85) discovered them. He gives
a photograph which shows boards of 4x 8, 4X9, and 4 Io cells.
4. Numerous boards cut in quartzite rock and much weathered on the Dwenyonyi
ridge to the south-west of Grant Bay on the Kyagwe coast of the Victoria
Nyanza, which appear to be associated with a line of fifteen stony mounds
(Wayland, b, 85).
5. Several boards near Umtali and Manica (Avelot, b, 14, citing v. Luschan, a, 879).

An exception to this lack of information as to date is supplied by some

rock boards in Ceylon to which Parker (pp. 224, 594) has directed attention.
At Pallebaedda, in front and a little to the right of the Wihara cave (dating
from the second century B.c.), is a board of 2x9 holes, about 14 in. in dia-
meter and from 14 to 1 in. apart, with a larger hole at one end. There are
also on the sloping surface of an immense rock, called Gaimaediyagala, ‘the
stone frog rock’, near the Siyamdalangamuwa tank (constructed between the
second and fourth centuries A.D.) no fewer than five sets of holes side by
side, each of 2x6 or 2X7 holes, two sets having larger holes at either end.
Parker suggests that these sets may have been made by the workmen engaged
in building the dam for the tank and used for playing mancala, but both the
position of the sets on the edge of a precipice and their arrangement with
very little room between the boards for the players rather tells against this
suggestion. No other, however, has been offered.
There remain the game-diagrams made on building-stones in ancient
Egypt to which I have already called attention. Parker (p. 594) gives the
following particulars:

Several well-cut sets of similar holes are to be seen on the roof-slabs of the Kurna
temple in Upper Egypt, and on the summit of the damaged portion of the great
 MANCALA. PART I 161
pylon built in Ptolemaic times at the entrance of the temple of Karnak, as well as on
the tops of the walls there and at the Luxor temple. The rows consist of six, seven,
and eight saucer-shaped holes on each side, the largest being 34 in. wide and 1 in. deep.
The finest set at Kurna has sixteen holes and is 2 ft. long; the holes, which are admir-
ably cut and finished, are 23 in. wide and 1 in. deep. In both these instances the centre
lines of the cups are from 5 to 6 in. apart. On the Karnak pylon I also saw some holes,
perhaps intended for this game, which consisted of two rows with only four and five
cups in each. In all these instances there are no surplus holes for the captured pieces.
Another set of twelve cups, 2 ft. long, consisting of two rows from 4 to 7 in. apart,
is cut at the south-east corner of the Pyramid of Menkaura at Gizeh on a large block
of the rock which has since fallen over on its side. The cups are from 2 to 2} in.
wide, and from 1 to 1} in. deep, with a large one at the side, 2} in. wide,
for holding captured pieces. These holes are much weather-worn, and are quite
different from the shallow saucers cut by the modern Arab guides at the Great Pyramid
of Khufu, for playing the game called Siga; and they have every appearance of a much
greater age. . . . It is possible that they were cut by the masons who were engaged in
the construction of the Pyramid or some of the tombs near it, since in Muhammadan
times there could be no reason why they should not be cut on the stones of the
Pyramid, like the modern ones, rather than on a distant part of the rough basal rock.

! 2 3 4 5 6
FIG. 76. MANCALA BOARD (Hyde)

In its simplest form, as used today in Egypt and Syria, the board for
mancala II is a piece of timber, about 2 in. thick and 6 in. wide, on one
surface of which two rows of holes about 2 in. in diameter are cut. There are
six holes in each row in Egypt, and seven in Syria,
but the games do not require any fixed number
of holes in each row; there may be as few as O
: 9 O00 0000
three in some children’s games, and as many as : :
fifty in the Masai game guehe (7.6.15). Different
peoples, however, usually show a preference for i
a particular number, generally six, seven, or
FIG. 77. PALLANKULI
eight in each row. The only difference produced 0 (resale Seathers
by increasing the number of holes is that it in- India)
volves an increase in the number of pieces, and
takes longer to bring the game to a conclusion. The Egyptian and Syrian
boards are sometimes made in two pieces, hinged together so that the board
may be closed like a box when not in use, and K. C. Murray tells me
that hinged boards are coming into use in Nigeria. The Tamils of southern
6366 M
162 MANCALA. PART I [CH 7
India also often use a hinged board, shaped to give the form of a fish when
it is closed.
In most countries, the board, when it is not hinged, contains, in addition to
the holes for play, two receptacles or stores which are usually placed one at
each end of the board, though they may be attached to the middle of each
side, placed between the rows, or both may be placed at the same end of the

FIG, 78. WARI BOARD, WEST AFRICA (Anthrop. Mus., Marischal College,
Aberdeen)

board. These stores are used for holding the pieces which each player may
capture in the course of a game. In Malaya, Indonesia, and the Philippines,
where the stores are placed at opposite ends of the board, each player includes

FIG. 79. AYO BOARD (B.M. 1928, 2-7, 1). Style of the Egba clan, Yoruba tribe,
Nigeria

his store among the holes in which he sows pieces. The board is often
supplied with feet or other means of raising it from the ground, and these,
particularly in West Africa, are often elaborately carved.
The general design of the boards for mancala III and IV differs from that
of the mancala II boards in having three and four rows of holes respectively.
They may, or may not, have stores, and some of the mancala IV boards
have oblong compartments instead of round holes. The boards are usually
without feet.
There is no discrimination of pieces necessary in any of the mancala games;
no special type of piece is required and all the pieces are shaped alike. Any
of the kinds of nonce-pieces given in Chapter 1 may be used, provided they
can be obtained in sufficient quantity. Perhaps beans, berries, or seeds are
most often used by inland peoples, or cowries near the coast. In the sequel I
shall call the pieces beans. A hole will be said to be loaded if it contains at least
two beans, even if it contains an even, and odd if an odd, number of beans. A
 MANCALA. PART I 163
single bean in a hole, and the hole in which it is placed, will be called a
singleton. Many games have rules limiting the power of moving from a
singleton.
All the mancala games are normally played by two persons whom I shall
call X and Y. They sit or squat on the ground facing one another with the
board placed lengthways between them. In all parts of Africa there is usually
a crowd of onlookers who discuss the moves, advise the players, and at times
even interfere with the play, so that it is easy for an uninstructed observer to
think that a number of persons is playing. It is convenient to regard each
player as owning the half of the board nearer to him. This is true of mancala
IV in which the moves of the beans on the player’s rows are strictly confined
to these rows, but only true of mancala II in the limited sense that each
player must begin a turn of play by moving the beans from a hole in his own
Tow.
Before play begins, the beans are usually arranged evenly in the holes, the
same number (four in most mancala II games or two in most mancala IV
games) being placed in each hole to ensure that the right number of beans is
used. Play begins from this arrangement in most games, but in many
mancala IV games play begins from a different arrangement and some preli-
minary moves have to be made to secure this arrangement; these moves will
be called the opening play. It is only in some Egyptian and Syrian games that
players are permitted to arrange their beans at random.
The players move alternately, and a turn of play is made up of one or
more laps. A lap includes the following operations: (1) the player lifts all the
beans (in a few games, all but one) contained in a hole on his side of the
board; (2) he deals or sows these beans one at a time in each of the consecutive
holes in the prescribed direction of movement. In most mancala games this
is anticlockwise round the board, but in some games it is clockwise or
determined by the player’s first move; in others it is alternately anticlock-
wise and clockwise. In mancala IV games it is anticlockwise round the
player’s two rows, but clockwise moves may be made in certain circum-
stances.
A lap ends when the last bean in hand is sown. What happens now depends
on the particular rules of the game that is being played, and these depend on
the situation of the hole in which the lap ends, its contents, and the contents
of the hole immediately opposite. The player’s turn of play may end with
or without captures, or, when the move is made up of several laps, the con-
tents of the hole in which a lap ends (or in some games the contents of the
hole next in front of it) are lifted for another lap which is played in the
same way.
For the rules of capture (which vary from game to game), the reader must
refer to the descriptions of the games below.
The result of a game depends on the number of beans taken by the two
players. In some games, the player who makes the greater number of
164 MANCALA. PART I [CH.7
captures wins. In others the game is won by the player who first succeeds in
making a given number of captures, and this may require a series of games
before this number is reached. In others the winner has to take all the beans
on the board and the game is played in rounds, each round beginning from a
new arrangement of the beans which a player holds at the end of the pre-
ceding round.
Mistakes, accidental or deliberate, may easily occur when sowing a handful
of beans one at a time. In two of the games which K. C. Murray recorded
for me, such mistakes happened accidentally.. But many observers have
borne witness to the prevalence of deliberate cheating. If this is detected, the
offender may be required to play the sowing again, or, less frequently, is
penalized by the loss of the game. In one game at least, the players are
required to sow from a hand well above the board to make mistakes easier
to detect. As a rule, as in archaic culture generally (Huizinga, 52), cheating
is not reprobated, and a successful cheat is rather admired, though to be
discovered entails a loss of prestige (Driberg, 114).
It will be obvious from this general description of the method of play
that the mancala games form a special class of board-games, and that they
do not exemplify any of the more ordinary activities of early man. I find it
difficult to believe that the mancala games can have been invented in vacuo,
and we seem to be driven back to the hypothesis that mancala arose out of
experimentation with an already existing board. But what purpose the man-
cala board may have served is not easy to see when no use is made of it
anywhere, now, except for a game. It may be significant that the earliest
boards all occur in the neighbourhood of building operations. May the
board have been used for the calculation of the wages to be paid to workmen,
and the board be originally a primitive kind of abacus?
MANCALA II
I have collected some 140 references to games on the mancala II board of
which about a hundred give more or less complete descriptions of the rules
of play. I arrange this material by the regions in which the games are
played, beginning with those played in Egypt and Syria, and proceed thence
eastwards to India, Ceylon, farther India, and the Far East, and then west-
wards and southwards to the games played in Africa. Whenever available,
I give the scores of games which have been actually played by natives of the
different regions, and adopt the following procedure:
The holes are described by letters, one row upper case (A, B, C, &c.), the
other lower case (a, b, c, &c.), and the lettering of each row starts from the
end left-hand hole as seen by the player who owns the row, and proceeds
anticlockwise along the rows.
The score is written in columns and the moves are unnumbered, as is
usual in books on draughts. Each player’s turns of play are preceded by X or
Y to show the player or followed by a line to separate the turns of play.
 MANCALA. PART I 165
Each lap is given a separate line.t Clockwise moves are preceded by an
asterisk (*).
A lap is defined by the number of beans lifted and the hole from which
they are lifted. Thus, X 4D means that X lifts four men from D and sows
them anticlockwise until they are all sown, and X*3C means that X lifts
three beans from C and sows them clockwise.
Captures follow the description of the lap and are preceded by a colon
(:), which is followed by the number of beans taken and the hole or holes
from which they are taken. Thus Y 7a: 4E means that Y lifts seven beans
from a and sows them, taking four beans from E.
The actual position after a particular move will sometimes be diagrammed
as a check when the score is played over on the board.

I. ARABIA, PALESTINE, SYRIA, TURKEY, IRAQ, AND IRAN

All these countries know the game by its Arabic name mancala, though in
central Arabia, as Doughty (i. 536) found at Teyma, it was also known collo-
quially as biatta (from Ar. beit, ‘house’; also used for the cell on other game-
boards). If several forms of the game are played, each may have its own
special name. The front hole on each player’s row is his ras, ‘head’. The bean
is called hasweh (pl. hasa), ‘pebble’ or kelb (pl. kilab), ‘dog’.
The earliest mention of mancala II in Arabic literature occurs in the Kitab
al-Aghani, compiled by Abii'l Faraj (A.D. 897-967), where reference is made
(Bulag ed., x1. xix. 6) to ‘a game resembling mancala’. We may note that,
unlike other board-games played by Arabs at that time, mancala has a purely
Arabic name, suggesting a native invention or at least a life of some duration;
and as mancala is not included in the list of games, the playing of which was
permitted, disapproved, or forbidden in the early Muslim codes of juris-
prudence, it looks as though mancala II was not widely played.
The first European traveller to notice the game in the Levant was Jean de
Thévenot (1633-67), who was there in 1657-9 and says in his Voyages, Paris
1665: ‘They play mancala very frequently, which is made in the shape of a
box, about two feet long and half a foot wide, with six small holes in the
box itself and six in the lid which is hinged to the box (for it opens like a
chessboard). Each player has 36 shells, six of which are placed in each hole
at the beginning of the game.’ Then, in 1694, Hyde (pp. 226-32) gave a
detailed description of two forms of mancala as played in Mesopotamia and
Jerusalem from information given him by native players. Unfortunately his
account of bagiira, the more intricate of these games, is confused. In my
description below, I have tried to unravel the tangle with the help of existing
games.
' | do this to help the reader who may desire to play the game over on a board. It would, of
course, have been possible to give a single line to each turn of play, but my own experience has
taught me that it is very easy to go wrong in playing a series of laps that make up a turn of play;
to give a line to each lap reveals such mistakes directly they are made.
166 MANCALA. PART I (CH. 7
According to Hyde, at the end of the seventeenth century, players in
Syria, Turkey, Mesopotamia, and Jerusalem used a 2x6 board while the
Arabs used a 2x7 board.-At the present time, the 2 x 6 board is only used in
Egypt, all the other countries having gone over to the 2x7 board. The
number of beans required for both boards is twice the square of the number
of holes in a row, seventy-two on the 2x6 and ninety-eight on the 2x7
board. All moves are made anticlockwise unless otherwise stated. Larger
boards are occasionally used; thus v. Luschan (Avelot, b) saw a 2X 12 board
with two stores in use at Aleppo, Asia Minor, but gave no particulars. The
ordinary boards in these countries have no stores.
7.1.1. Mesopotamia: Haliisa, haliisi (Hyde, 226). 2x6 holes; six beans
in each hole; one lap to the move; one round; moves anticlockwise.
A move can start from any hole on the player’s row, and if sowing ends
in a hole on his row, making it even, its contents are taken. If this hole is
preceded by an unbroken sequence of even holes, their contents are also
taken. If either player on completing his move has all his holes empty, his
opponent must, if possible, make a move which leaves him something to
play with. The game ends when either player is unable to move; his oppo-
nent then takes all the beans left on the board, and the player wins who has
the larger number of beans.
7.1.2. Island of Hydra off the east coast of Greece: Mandoli, ‘almonds’
(Galt, 242). Probably the same game as 7.1.1, but Galt does not say that the
hole from which captures are made must lie on the player’s own row. He
adds that neither player may make a capture on his first move, but this
seems impossible in any case.
7.1.3. Mesopotamia: Baqitira (Hyde, 230). 2x6 holes; each player has
thirty-six beans which he arranges as he likes in his holes, but usually leaving
both end-holes empty; several laps to the move; several rounds; moves
anticlockwise. A singleton cannot be moved so long as the player has any
loaded holes. A new lap begins, not from the hole in which the previous lap
ended, but from the next hole in advance.
A move may begin from any hole on the player’s row. If sowing ends in
a hole on his own row making it even, he takes the contents of this and the
opposite hole, and if the hole is preceded by an unbroken sequence of even
holes, he takes their contents and the contents of the opposite holes. If sowing
ends in a hole on his own row making it odd, or in a loaded hole on the
opponent’s row, the lap ends and the player begins a new lap. If sowing ends
in an empty hole, the move ends. The round ends when one player is unable
to move.
For a new round, the player who now has the smaller number of beans in
his store arranges his beans as he likes in his holes, and his opponent copies
this arrangement in his own holes, leaving any surplus beans in his store.
The round is then played as described above, and the game continues with
further rounds until one player has lost all his beans.
7.1.4] MANCALA. PART I 167

7.1.4. Egypt: Li‘b al-ghashim, ‘game of the unlearned’ (Lane, 315). 2x6
holes; seventy-two beans which one player distributes unevenly in the holes
on both rows, putting at least four in each hole except the end-holes which
usually are left empty.’ His opponent then moves first, but if he is dissatisfied
with the arrangement in his holes, he may turn the board round, but then
the dealer begins. A series of games is played for sixty points, and after
each game the player with the larger number of beans scores the number
he has in excess of his opponent. Several laps to the move; play anticlock-
wise.
Every move must begin from the player’s ras or from the nearest loaded
hole behind it. If a lap ends in an empty hole, the move ends; if in a hole
making it even, the player takes its contents and the contents of the opposite
hole, and if the hole is preceded by an unbroken series of even holes, he also
takes their contents and those of the opposite holes. If the lap ends in a hole,
making its contents odd, the lap ends and a new lap is begun from this hole.
If a player whose turn it is to play has no beans in his holes, his opponent
must give him a bean, placing it in the player’s left-hand hole so that he may
be able to play. When the board is cleared, the game ends, the score is
calculated, and the seventy-two beans are arranged for a new game, and
further games follow until one player’s score reaches sixty.
7.1.5. Damascus: Li*b majniini, ‘the crazy game’ (Culin, c, 597). 2X7
holes; ninety-eight beans which one player distributes unevenly in the holes
as in 7.1.4, except that at least two beans must be placed in each hole. Several
laps to the move; one round; play anticlockwise.
All moves must begin from the player’s ras or the nearest loaded hole
behind it. If a lap ends in an empty hole, the move ends; if in a hole, making
its contents two or four beans, these and the contents of the opposite hole
are taken, and if it is preceded by an unbroken sequence of holes containing
two or four beans, these and the contents of the opposite holes are also taken.
Otherwise, the contents of the hole are lifted for a new lap. The game ends
when one player can no longer move, and the player wins who has made
the larger number of captures.
7.1.6. Damascus: Li‘b roseya (Culin, c, 598). Played by children. 27
holes; seven beans in each hole.
7.1.7. Egypt: Li‘b al-‘aqil, ‘game of the learned’ (Lane, 317). 2x6 holes;
seventy-two beans. Only differs from 7.1.4 in greater freedom in the arrange-
ment of the beans—thus a player may put about half in A and the remainder
in a—and in allowing the players to count the number of beans in the holes
after the first move, and to see and count the number of beans lifted by the
opponent for sowing to ensure that they are sown correctly.
7.1.8. Damascus: Li*b hakimi or “aqila, ‘rational or intelligent game’
(Culin, ¢, 598). 2x7 holes; seven beans in each hole. Only differs from 7.1.5
T Lane says the beans are arranged unevenly because, if they were arranged evenly, the dealer
would be sure to win.
168 MANCALA. PART I (Coa,
in that moves may be made from any hole on the player’s row and the
players may count the contents of the cells.
7.1.9. Bedouins, Arabia: Mangala (Parker, 601). 2x6 holes; seventy
beans which one player distributes unevenly in C, D, c and d, leaving two
empty holes at each end of each row. The opponent then begins play, but
first with closed fist feels the contents of the holes and if he is dissatisfied with
the distribution, he can turn the board round and the dealer begins. Several
laps to the move; one round; moves clockwise. The first move must be
from D (d) and no capture can be made on this move. Subsequent moves
may start from any hole on the player’s row.
If the last bean of a sowing falls in a heavily loaded cell, its contents are
lifted for a new lap; if the hole contains not more than ten beans and the
number is odd, the move ends; and if even and the opposite hole is also
even, the contents of both are taken, otherwise the beans are lifted for a
fresh lap.
7.1.10. Suez, Egypt: Mangala (Parker, 601). 2x6 holes; two holes in one
row and one in the other are left empty, and eight beans (cowries, called
kilab, ‘dogs’) are placed in the remaining nine holes. Several laps to the move;
one round; moves anticlockwise. The first move of each player is made
from any of his holes, and laps are played until the last bean in hand falls in
an empty hole, when the turn of play ends; no captures may be made in
these opening moves. Thenceforward, each player begins his move from the
hole next beyond the hole in which his opponent finished or the first loaded
hole beyond it. The move continues by laps until the last bean in hand falls
in an empty hole. If the opposite hole contains beans, these are taken, and
the player transfers the last bean he has sown to the next hole and lifts the
beans in it for a new lap, and continues until the last bean in hand falls in a
hole opposite to an empty hole, when the turn of play ends. Each player
must always begin his turn of play after the first moves from the next hole
to that in which the previous turn of play ended, or the first loaded hole
beyond it.
7.1.11. Teyma, Arabia: Beatta, ‘holes’ (Doughty, i. 536). 2x7 holes;
seven beans (gaud, ‘camel foals’) in each hole.

2. INDIA

Mancala II is played in most parts of India, but, though it is said to be

popular, very few observers have troubled to master the method of play.
It is not included in Herklots’ list of games in the Deccan, but Culin (c, 600)
was told that it was common in Bombay, and Mr. A. G. Shirreff tells me
that he had seen it played in the Punjab and United Provinces and by children
in the suburbs of Lucknow. In northern and central India there are generally
five beans in each hole, and when the move consists of a number of laps, the
beans in the hole in front of the hole in which the previous lap ended are
lifted for the new lap.
7.2.1] MANCALA. PART I 169
7.2.1. India, Punjab: Khutka boia (H. C. D. Gupta, d, 211). 2x5 holes;
five beans in each hole.
7.2.2. India, United Provinces, Karwi Subdivision: Pachgarhwa (E. de
M. Humphries, 117). 2 5 holes; five beans in each hole. Several laps to the
move. When the last bean in hand is sown, the contents of the hole imme-
diately in front of this are lifted for the new lap. If this hole is empty, the
player captures the beans in the hole immediately in front of it. The account
goes on to say ‘later things begin to get complicated, and I was unable to
discover how, if ever, the game did end’.
7.2.3. Assam: Mawkar katiya: (H. C. D. Gupta, b, 71). At Khasis,
Cherrapunji (H. C. D. Gupta, b, 71-74). 2x7 holes; five beans in each
hole. Several laps to the move. The move ends when the player comes to
an empty hole when he takes all the pieces in the next hole.
7.2.4. Lushei Kuki clans, Assam: Vai lung thlan, ‘foreign stone gravel’
(Shakespear, 39). 2x6 holes; five beans in each hole; one lap to the move;
one round; moves clockwise; played by men and women.
A move may begin from any hole on the player’s row. If the last bean in
hand is sown in an empty hole, this is captured and if the hole is preceded by
an unbroken series of singletons, these are also taken. Deliberate counting of
the beans in a hole is forbidden.
7.2.5. Lakhar tribe, Assam: Longbeu-a-cha (N. Parry, 189). 2x 5 holes;
five beans in each hole; several laps to the move; moves anticlockwise.
A move may begin from any hole on the player’s row. When the last
bean in hand is sown, the contents of the next hole are lifted for a new lap.
If this hole is empty, the move ends but the contents of the hole beyond are
taken. If the holes beyond are alternately empty and loaded, the contents of
the loaded holes are also taken.
7.2.6. Orissa: Kanj guti (H. C. D. Gupta, b, 71).
7.2.7. Chota-Nagpur, Mudas, and Birhots tribes: Til guti (CP.). 2x7
holes and two stores, one at each end of the board; five beans in each hole.
7.2.8. Madras: name unrecorded (H. C. D. Gupta, b, 71).
7.2.9. Tamils, southern India: Pallankuli, ‘many holes’ (Mrs. H. G.
Durai, 135). Played by women only. 2X7 holes; occasionally stores, but the
fish-shaped boards have none; six beans in each hole; several laps to the
move; several rounds; moves anticlockwise.
A move may begin from any hole on the player’s row, even though this
is a singleton. If in the course of sowing, the contents of any hole are brought
up to four, these are taken and sowing continues until the last bean in hand
is sown. If this falls in an empty hole, the contents of the next hole are taken;
if in a loaded hole, the contents of the next hole are lifted for a new lap. The
round ends when one player has no beans left on the board, and the opponent
takes any beans in her own holes.
For the new round, the player who has taken the smaller number of beans
fills as many of her holes as she can, starting from her left-hand end-hole
170 MANCALA. PART I (CH. 7
and putting six beans in each hole. Any surplus beans that she has, she returns
to her store. The other player then does the same, retaining any surplus beans
in her store. Empty holes are now “dung holes’ and are out of play for the
new round. Further rounds are played until one player is reduced to fewer
than six beans. The right to begin alternates from round to round.
7.2.10. Maldive Islands: Naranj (Culin, c, 600). 2x8 holes and two
stores, one at each end of the board; five beans in each hole.
7.2.11. India, Central Provinces: Sat-gol (H. C. D. Gupta, c, 165). This
game belongs to the mancala class, but differs in the form of the board.
Seven holes arranged as a circle; four beans in each hole. Several laps to the
move. New laps begin from the next hole to that in which the preceding lap
ended. Move ends if this hole is empty and the player captures the contents
of the hole beyond, or if there is a sequence of empty holes, the contents of
the first loaded hole.
3. CEYLON
Mancala is known in Kandy and the interior of Ceylon as olinda keliya,
‘the olinda game’, from the seeds of the olinda creeper which are used as
beans. It is almost exclusively played by women, and usually at the season
of the New Year. Parker describes fully four varieties, all of which are
played on the board of 27 holes, though there is evidence (see p. 160) that
formerly other numbers of holes were used. All the native games are played
in rounds and occupy a very long time.
In Colombo and the coastal areas where there is a considerable number of
Tamils and Muslims, a different form of game is played with cowries for
beans, which is known as chonka, chonku, or chanka from the Malays, or
pallankuli from the Tamils. Parker notes that the holes in the Colombo
boards are larger than in the Kandy ones, probably because cowries are
larger than the olinda seeds. Both boards have two stores (tachi), rectangular
or oval in shape and larger than the holes, which are placed sometimes at
the ends, sometimes at the sides, and sometimes between the rows of the
board.
The rows are called pila; the holes wala, ‘hole’ or kotuwa, ‘enclosure’; an
empty hole is puhuwala or pussa; the bean is a ‘dog’ in the coastal regions and
indiya (pl. indi), ‘date’ in the interior; to sow is ihinawa, ‘to sow’; to capture
is kanawad, ‘to eat’.
7-3-1. Kandy, Ceylon: Puhulmuti, ‘ash-pumpkin pearls’ (Parker, 594).
2x7 holes and two stores; four beans in each hole; several laps to the move;
several rounds. All moves are made in the same direction, all clockwise or all
anticlockwise. The player who begins usually lifts the beans from the second
hole from either end of the row and sows the first bean in the adjoining end-
hole, thereby fixing the direction of play for the round. Subsequent moves
may be begun from any of the player’s holes. Singletons cannot be moved if
the player has any loaded hole, and a singleton in her front end-hole cannot
7.3.1] MANCALA. PART I 171
be moved if she possesses other singletons. When sowing beans, none can be
sown in a hole which already contains three beans, except the last bean in
hand, in which case the four beans which it now contains are taken, and the
contents of the next hole in front are lifted for a new lap or, if it is empty,
the move ends. Otherwise, if the last bean falls in a loaded hole, its contents

ex
C2.
Olinda seeds

Colombo board
FIG. 80. CEYLON BOARDS (Parker)

are lifted for a new lap, and if in an empty hole, the move ends. The round
ends when one player has all her holes empty.
Second round. The winner of the first round puts four beans in each of her
holes, leaving any surplus beans in her store. The loser then, starting from
one end of her row, puts four beans in as many of her holes as she can and
leaves any beans over in her store. The holes which she cannot fill become
kana, ‘blind’ (also used for the player herself) and are out of play for the
round, It is usual to place a twig or straw across the blind holes to distinguish
them. The blind player begins the round, moving towards her blind holes,
and the round is played in the same way as the first round.
172 MANCALA. PART I [CH.7
Third and later rounds. The winner of the previous round puts four beans
in each of her holes, and the loser puts four beans in as many of her holes,
starting from one end, as she can, and the odd beans over in the next hole.
If this hole contains one bean, it is called puta, ‘son’; if two, naga, ‘younger
sister’; if three, wala, ‘slave’. The holes with no beans become blind and are
out of play for the round. If the loser has a puta hole, the opponent removes
one bean from her opposite hole; if the loser has a naga hole, the opponent
removes two beans from the opposite hole; and if she has a wala hole, the
opponent removes three beans from her opposite hole, making one of her
holes wald, naga, or putd respectively. She puts the beans which she has
removed in her store. The putd and naga holes are then marked by putting a
piece of paper or straw in them. The blind player then begins, playing in the
direction of her blind holes and so fixing the direction of move for the round.
Beans are sown in all the holes except the blind ones, but beans in putd and
naga holes cannot be lifted for sowing or taken; they accumulate for the
player on whose row they lie. If the last bean in hand falls in a puta or
naga hole, the move ends. The round is played in the same way as the
first round.
When one player is left at the end of a round with fewer than twelve holes,
she has the option of arranging them differently for the next round. She may
put one or two beans in one end-hole and not more than four in the other,
and one or two beans in some of the intermediate holes, leaving the others
blind. Thus, with nine beans, she may arrange them 2, 2, 2, 2, 0, 0, 1. Her
opponent then puts four beans in each of her holes. There are no puta, naga,
or wala holes in this round, and the players play in different ways. The
player with full holes plays as in previous rounds, but the blind player only
takes when her last bean in hand falls in a hole on either row, making its
contents three beans if she had put two beans in her first end-hole, or making
its contents two beans if she had put one bean in her first end-hole. Other-
wise, except when sowing the last bean in hand, she omits all holes containing
one or two beans, but not those which contain three beans as in the first
round. In this way the blind player may regain her lost beans and the game
may run on for hours.
In order to shorten the game, the blind player may play puhul kapanawa,
‘cutting ash-pumpkins’. In this she borrows a bean from each of her oppo-
nent’s last two holes and places them in. her opposite holes. She must then
begin her move from her third hole and return the borrowed beans after
the end of her move. Or she may move one or two beans on her opponent’s
row and begin her move from other holes (so Parker).
7-3-2. Kandy, Ceylon: Walak-pussa, ‘a hole empty’ (Parker, 597). 27
holes and two stores; four beans in each hole; several laps to the move;
several rounds. The game is begun like 7.3.1, but laps are played and captures
made differently. When a lap ends, the beans are lifted from the next hole
for a new lap. If this hole is empty and the one in front of it is loaded, the
7.3.2] MANCALA. PART I 173
beans in it are taken, and so long as holes are alternately empty and filled,
the beans in the filled holes are taken, and the move ends (cf. 7.2.3).
Further rounds are played as in 7.3.1, but with puta, naga, and wala holes.
If a lap ends in a hole preceding these holes, they are treated as empty and
the contents of the next hole beyond are taken.
7.3.3. Kandy, Ceylon: Kotu-baendum, ‘tying up the holes’ (Parker,
598). 2X7 holes and two stores; four beans in each hole. Played like 7.3.1.
but the player’s first move must begin from one of her end-holes; subsequent
moves may begin from any of her holes. If a lap ends in an intermediate
hole containing three beans before the bean is sown in it, the four beans are
taken; if, however, it is sown in an end-hole containing three beans, the
hole is tied (baenda) and treated as a putd or naga hole, and its contents are left
to accumulate for the benefit of the player who made it tied. If the last bean
in hand falls in a tied hole, the move ends.
7.3-4. Kandy, Ceylon: Daramutu or Ellaewala-kanda (Parker, 599).
27 holes and two stores; four beans in each hole. Played like 7.3.1, except
that play may begin from any of the player’s holes, and if a lap ends in an
empty or puta or naga hole, the contents of the opposite hole are taken and
the move ends.
7-3-5. Colombo, Ceylon: Pallankuli or Chonku or Chonka (Parker,
$99). 2X7 holes and two stores; seven beans in each hole; several laps to
the move; one round.
The player who begins makes her first move from any of her holes, and
the direction in which she makes it determines whether moves shall be clock-
wise or anticlockwise. When sowing, each player includes her store in the
holes in which she sows beans and sows in it in its turn. If the last bean in
hand falls in an empty hole, it and the contents of the opposite hole are taken
and the turn of play ends; if it falls in the player’s store, the turn of play
also ends; if it falls in a loaded hole, the contents of this hole are lifted for a
new lap.
4. SIAM, ANNAM, SOUTHERN CHINA, MALAYA, INDONESIA, AND
THE PHILIPPINE ISLANDS
The outstanding feature of the mancala games of these countries is the
inclusion of each player’s store among the holes in which he sows beans.
Sir G. G. Scott (Shway Yoe) does not include mancala in his list of games
played in Burma.
7.4.1. Siam: Mak khom (Capt. Low, 380). 2X7 holes; seven beans in
each hole. Capt. Low’s account is rather sketchy, so I quote it.
Each takes out the contents of his first cup and counts them out to the right hand to
the last number, setting aside the counter that remains. They then begin with the
second one, and, when the contents have been told out, they respectively take out of
the cups (amongst which the last told out counter falls) their contents, and proceed to
count out as before. They may take the whole of the numbers in any cup or only part
174 MANCALA., PART I [CH.7
of it. The partners agree that after a certain number has been won by one of them,
the game shall cease, as it may be immoderately lengthened out.
I infer that play is anticlockwise and that the player sows in his own store,
but nothing is said about the way in which captures are made.
7.4.2. Annam: name unrecorded (Ch. Beurt, ‘A propos du Jeu de |’Awele’
in Notes Africaines, Jan. 1947, 13). 2X $ holes with two stores, one at each end
of the board. Beurt refers to a painting by an Annamese artist, N’Guyen
Phan Chanh, ‘Les ébats des enfants’, which shows Annamese children playing
a game which Beurt identified with the West African game wari.
7.4.3. Southern China: name unrecorded (ENI.).
7.4.4. Malaya: Main chongkak (Skeat, 485; Wilkinson, 57; Culin, ¢,
600; M. Hellier, 93; H. Overbeck, 8). 2x6 holes (Wilkinson); 2x7 holes
and seven beans in each hole (Hellier); 2x 8 holes (Skeat; Culin); 2x 9 holes
and nine beans in each hole (Overbeck). Play is always clockwise, and each
player owns the store to his left and includes it when sowing beans. Several
laps to the move; several rounds. Played by women.
A move may begin from any hole on the player’s own row. If in sowing,
the last bean in hand falls in an empty hole on her own row, she takes the
contents of the opposite hole and the move ends; if in her store, she begins
a new lap from any of her holes; if in an empty hole on her opponent’s row,
the move ends; if in a loaded hole on either row, the contents of that hole
are lifted for a new lap. The round ends when either player’s holes are all
empty, and the player who has beans in her holes adds them to her store.
The beans are then rearranged in the holes, each player using the beans in
her store and putting in each hole from right to left the same number of
beans that they contained when the game began, and returning any surplus
beans to her store. Any unfilled holes are ‘dead’ and out of play for the new
round. The round is played in the same way as the first round, and further
rounds are played until one player has too few beans left to fill a single
hole.
7.4.5. Sumatra: Chato (ENI.), the generic name for any mancala game
in northern Sumatra. Also chuka, jungka. All these words are used loosely
for other board-games. At Achehn four varieties are played (Snouck-Hur-
gronje, ii. 200), meusuéb, meuta’, meuchoh, and meuliéh, but only the
first is described.
7.4.6. Achehn, Sumatra: Meusuéb (Snouck-Hurgonje, ii. 200). 2x6
holes (rumoh) and two stores (geudong or choh), one at each end of the board;
four beans in each hole; several laps to the move; one round; play anti-
clockwise. Much played by women and children.
A move may begin from any of the player’s holes. If the last bean in hand
falls in an empty hole, the move ends; if in a loaded hole, its contents are
lifted for a new lap, but if the hole now contains four beans, these are taken
and the new lap begins from the next hole. When both players have too few
beans left to play with, each plays one move with one of the beans left in
7.4.6] MANCALA. PART I 175
her holes, and then each takes the beans left in her holes, adds them to her
store, and the game is over.
Wooden boards are sometimes used, but as a rule the board is marked out
on the ground.
7.4.7. Central Sumatra: Bajangkaq, Kaloleh (ENI.).
7.4.8. Java: Dakon (Raffles, i. 250; D. M. Campbell, ii. 1017; ENI.;
MVB., Java, 3e Teil, 1349-5). 2x7 holes and two stores, one at each end of
the board; played with fourteen or eighteen beans (Raffles). A women’s
game.
7.4.9. Bali: Mechiwa (ENI.). Two boards in the Munich Museum.
7.4.10. Eastern Soemba: Matoe (Leiden Museum). The board is said to
have thirteen ‘depressions’.
7.4.11. Flores: Sai (S. Jacobsen in MVB., vi. i. C. 17863, 462). 2 7 holes;
four beans in each hole. A move can begin from any of the player’s holes,
and if it ends in a hole making its contents four beans these are taken. When
all the beans are taken, the loser pays the difference between his and the
winner’s captures.
7.4.12. West coast of Borneo: Chonka (ENI.).
7.4.13. Periching tribes, Central Borneo: Aw-li on-nam ot-tjin, ot-jin,
‘fish’ (C. Lumholtz, ii. 435). 2 10 holes and two stores, one at each end of
the board; from two to five beans, but usually three, in each hole; several
laps to the move; one round; play anticlockwise.
A move may begin from any of the player’s holes. If the last bean in hand
falls in an empty hole, the move ends; if in a hole making its contents the
same number of beans as that originally placed in the holes, these are taken;
if in a hole making its contents any other number of beans, they are lifted for
a new lap.
7.4.14. Celebes: Galatjang (W. Kaudern, iv. 313, quoting Matthes,
Macassar Dict., pp. 84, 898). 27 holes (kalobang) and two stores (anrong),
one at each end of the board, each player owning and sowing in the store
to his right hand; seven beans in each hole; several laps to the move; several
rounds; play anticlockwise.
A move may begin from any of the player’s holes (excluding his store).
If the last bean in hand falls in his store, he begins a new lap from any of his
holes; if in an empty hole on his own row, the move ends, but he takes the
bean he has just sown and the contents of the opposite hole; if in a loaded
hole, its contents are lifted for a new lap.
When no more moves can be made, the round ends and the beans are
arranged for a new lap, each player using his beans to fill as many of his
holes as he can by putting seven in each. Surplus beans are returned to his
store and any unfilled holes are said to be ‘burnt’, and are out of play for the
new round,
Matthes says that it is customary to play this game at the time of
mourning.
176 MANCALA. PART I [CH. 7
7.4.15. Macassar Peninsula, Celebes: Dara-dara (Kaudern, 317). 2X6 or
27 holes and two stores, one at each end of the board. Four beans in each
hole.
7-4.16. Philippines: Chuncajon (Culin, f, 654). 2x7 holes and two stores,
one at each end of the board, each player owning and sowing in the store
to his left hand; seven beans in each hole; several laps to the move; one
round; play anticlockwise.
A move can begin from any of the player’s holes. If the last bean in hand
falls in his store, the move ends. The description is incomplete as Culin says
nothing about captures.
7-4.17. Iloko tribe, Mountain Province, Philippines: Chuncajon or
Agsinnoninka (W. Vanoverbergh, 216). 2x 5 holes and two stores, one at
each end of the board, each player owning and sowing in the store to her
right hand; the same number of beans in each hole; several laps to the move;
several rounds; play anticlockwise. Played by young girls.
A move may begin from any hole (including singletons) in the player’s
row. The two players move simultaneously, beginning each move together.
When the last bean in hand is sown, the contents of the next hole are lifted
for a new lap. If the next hole is empty, the move ends. If the lap ends in one
of the player’s holes, the contents of the opposite hole are taken. When no
further moves can be made, the round ends.
Each player then redistributes the beans from her store, putting in each
of her holes the same number of beans as in the initial arrangement, as far
as is possible, and returning any surplus beans to her store. Any holes that
are not filled are said to be ‘burnt’ and are out of play for the new round.
When one player has too few beans to fill a single hole, the game ends.
5. WEST AFRICA; GUINEA FROM THE SENEGAL TO THE GABON,
AND THE SUDAN

F. v. Luschan (Die Altertiimer von Benin, 1918, p. 427) gives particulars of

five ancient bronze boards for games of wari, which were looted from Benin,
Nigeria, after the capture of the town in February 1897, and were in Euro-
pean collections in 1918. Three of these boards are certainly not mancala
boards, but there is some evidence that wari is occasionally used generically
for other West African board-games (cf. A. W. Cardinall, 255).! The remain-
ing two boards are unmistakably mancala boards with 2x 5 holes and cen-
trally placed stores, the whole standing on a pedestal (see Fig. 81). Similarly
arranged boards are still used in Benin. See 7.5.53 and $4.
The earliest mention of a mancala game in West Africa appears to be that
1 These boards are rectangular in shape and are raised on low feet; the surface is marked with
a large number of small holes, such as might be used for pegged men, arranged lengthwise in a
number of rows which are separated by lines. Another bronze board from Benin was offered for
sale by W. D. Webster of Bicester in 1899 (IIlustrated Catalogue, no. 21, item 219). This board is
divided into two squares with their diagonals drawn, which are separated by a rectangle of the
same width. The purpose of these boards is unknown.
7.4.17] MANCALA, PART I 177
in Richard Jobson’s Golden Trade, 1623 (Reprint, p. 48) in which he describes
his ascent of the River Gambia in 1620. He says:
In the heat of the day, the men will come forth and sit themselves in companies,
under the shady trees, to receive the fresh aire, and there passe the time in communica-
tion, having only one kind of game to recreate themselves withall, and that is a peece
of wood, certaine great holes cut, which they set upon the ground betwixt two of
them, and with a number of some thirty pibble stones, after a manner of counting,
they take one from the other, untill one is possessed of all, whereat some of them are
wondrous nimble.

FIG, 81, BRONZE BOARD FROM BENIN (Rushmore Collection. After v. Luschan)

Mancala is not included in the list of games played in the kingdom of

Ouidah which is given in R. P. Labat, Voyage du Chevalier des Marchais en
Guinée, 1730, ii. 215-20, but is said by P. E. Isert, Voyages en Guinée, trad. de
Pallem., 1783, p. 205, to be widely played by the Accra of the Gold Coast on
a board with fourteen deep cups, while Golberry, Fragm. d’un Voyage en
Afrique, 1791, ii. 480, says that mancala is played by the Peulhs, Joloffs, and
Mandigos, who call it ouri. T. E. Bowdich, Mission from Cape Coast Castle to
Ashantee, 1819, notes that the Ashanti played mancala under the name of
vora, and that their king had a board made of gold. B. C. Monrad, Gemalde
der Kiiste von Guinea, 1824, says that the Accra women on the Gold Coast
played it under the name of tami. These references, taken from Culin (c),
Avelot (a, 6), and Parker, are the only ones before 1830 that I know, but
there is a fine board from West Africa (see Fig. 78) in the Anthropological
Museum, Marischal College, Aberdeen (length 690, breadth 130, and height
III mm.), which was presented to the museum by Sir James McGrigor in
1820. The game was called kyngee bel or keya or wari-wari.
5856 N
178 MANCALA. PART I [CH.7
The games included in this section are for the most part of a simpler type
than those played in East Africa. The number of holes in each row ranges
from three in some children’s games to twenty in the okwe or azigo of the
Cross River district which has spread westwards to the Ibos and other tribes.
The area from which the games are drawn is bounded on the south by the
fringes of the Equatorial Forest Zone and on the east roughly by the meridian
20° E., but I also include some games from the Anglo-Egyptian Sudan,
which are identical or show marked affinities with West African games. We
have many accounts of games played to the south of the Sahara, to which
K. C. Murray has made a number of additions, but we know very little
about the games played in Morocco, Algeria, Tunisia, and Libya; indeed,
the only mention that I have found is that made by Lieut. Avelot (b, 11) who
was told by Dr. Hamy that he once came across some Berbers in southern
Tunisia who were playing a game with pellets of goat’s dung in small holes
in the sand, but the players tookto flight when he approached.
Mancala may have reached West Africa by two routes which have been
followed by man from early times, one north of the Sahara along the
Mediterranean littoral and then across the desert, the other south of the
Sahara from the southern end of the Red Sea to Calabar and the Cameroons,
but with the lack of any information as to the age of the game in West
Africa, nothing definite can be said. I am inclined to think that the 2x6
board, on which the majority of the West African games are played, arrived
by the northern route, while the games on longer boards came by the
southern route.
Various generic terms are used for mancala in West Africa. The most widely
spread is wari! which is current from Senegal to the Gold Coast. Adi (adji,
adjito) is used by the Ewes in Dahomey; ayo by the Yorubas of Dahomey
and Nigeria; okwe and azigo by the Ibos, Umons, Ibibios, and Owerris;
darra by the Hausas in Nigeria; i by the Dogons (about 150 miles south-south-
west of Timbuktu); songo in the Cameroons and Gabon. Some of these
names are borrowed from the name of the board; others from the seeds or
beans with which the game is played. Local varieties often have their own
distinctive name.
I arrange the recorded games under the following types:
When lifting beans from a hole for sowing, all the beans are lifted.
A. The move ends in a hole on the player’s row. The contents of the opposite
hole are taken.
B. The move ends in a hole on the opponent’s row making its contents a given
number of beans (generally two or three). These and the contents of an un-
broken sequence of holes on the same row with the same number of beans
are taken.
C. The move ends in a hole on either row, making its contents four beans. These
t Variously transliterated by observers as aware, woro, oura, awele, walu, wolo, and from the
West Indies, wawee.
7.5.1] MANCALA. PART I 179
are taken. If in sowing any hole contains four beans exactly, each player
takes these fours when in his own holes.
When lifting beans for sowing, one bean is left in the hole.
D. The move ends in a hole on the player’s row and the opposite hole contains
one, three, and sometimes five beans. These and the contents of an unbroken
sequence of holes similarly loaded are taken.
E. Only differs from D in that clockwise moves may be made, provided they
result in a capture.
F, The game begins with ten beans in each hole.
Games with different directions of sowing.
G. The move ends in a hole making its contents two or four beans.
H. Only differs from G in that moves may be made clockwise and anticlock-
wise.
I and J. Games which do not fall in any of the preceding types.

Before dealing with the mancala games proper which are played in this
area, I give some games or puzzles in which a different use is made of the
mancala board.
7.5.1. Senegal (Golberry). This is a variety of the arithmetical puzzle
which Kraitchik (83) describes under the title “The Battle of Numbers’. I
give Golberry’s account from Avelot, a:

Chaque joueur a 21 boules 4 placer; les coups sont alternatifs, et la primauté se tire
au sort. Il faut qu’a la fin de la partie, les six cases de chaque cété aient été employées;
qu’a la sixiéme case, les 21 boules aient été toutes placées; et, par la combinaison de ce
jeu, le gain de la partie appartient 4 celui qui a mis son adversaire dans l’impossibilité
de jouer son dernier coup, en observant les régles du jeu qui varient suivant certaines
circonstances de la partie, et suivant le nombre pair et impair des premiéres boules
placées. Par exemple, si celui qui a gagné la primauté débute par ne placer qu’une
boule, son adversaire est forcé par la régle du jeu, d’en placer au moins deux; mais il
peut aussi en placer trois. Celui qui a joué le premier, et qui va jouer son second coup,
est forcé de placer un nombre pair de boules, soit au-dessus du nombre trois, soit au-
dessous. L’intrigue et la combinaison de ce jeu consistent 4 mener la partie de maniére
qu’au dernier coup 4 jouer il soit resté dans la main de ]’adversaire un nombre de
boules qui ne soit conformé 4 la circonstances de la partie, et aux régles du jeu. Si cela
arrive 4 celui qui a eu la primauté, il a perdu, et l’adversaire est dispensé de jouer son
dernier coup; si cela arrive 4 |’adversaire, la partie est perdue pour lui; mais, si celui-ci
a joué de maniére qu’il arrive aussi 4 son dernier coup avec un nombre de boules
conformé aux conditions et aux régles du jeu, la partie est nulle. L’ouri a plus de
combinaisons que le jeu de dames; et cependant les femmes seules le jouent, et les
hommes ne s’en amusent jamais.

7.5.2. Sudan: Li‘bat Iblis, ‘the Devil’s game’ (Davies, SNR., viii.
148). 2x4 holes; four beans in each hole; two players; play anticlock-
wise.
X removes two beans from A and Y takes the remaining two beans in A.
180 MANCALA. PART I [CH.7
This is repeated with holes B, C, D, a, b, c. X then takes either one or three
beans from d and Y takes the remaining three or one. The players then put
back the beans, Y beginning by putting four beans in A, and X then puts
four in B. This replacement is repeated with holes C, D, a, b, c. When the
players come to put the beans they hold in d, they find that the number of
beans that they hold have been changed over from the number each took
from d.
7.5.3. Dogon tribe, French West Africa; Sanga: Gamma, ‘cat’; Ibi.:
Nyuwunu bo(uh) ana ya(i), “The dead play at ana’ (Griaule, b, 170). 2 4
holes. The same trick as 7.5.2.
7.5.4. Benin, Nigeria: “The magical game or contest between Aruanran
and Esigie, two of the sons of Oba Ozolua’ (J. U. Egharevba, 12). 2x 6 holes;
four beans in each hole. The same trick as 7.5.2. When the last hole (d) is
emptied, one player taking one bean and the other three, the players hold
their beans in the closed fist and shake them as they say ‘Aruanran’s beans are
to return to Esigie and Esigie’s to Aruanran, change, change, change’, and
return the beans to the holes; and when they come to hole d, they find the
number of beans has changed over.
7.5.5. Sudan: El arneb, ‘the hare’ (Davies, SNR., viii. 149). 2X3 holes
and two stores, one at each end of the board; fourteen beans, arranged, three
in the left-hand store and one in the other store, two beans in A, C, a, c, and
one in B and b; play anticlockwise. Played with laps. The player begins by
lifting two beans from c and sows them using both stores and holes. Sowing
never comes to an end as the original arrangement repeats after twenty-six
circuits of the board.
7.5.6. Benin, Nigeria: Ise~-Ozin-egbe, “Perseverance game’ (Egharevba,
26). 2X 3 holes and no stores; the beans are arranged, one in A and C, ten in
B and b; one player; played by laps; play anticlockwise. A similar game to
7.5.5. The player begins by lifting the ten beans in b and the move never
ends, the initial arrangement recurring (according to Egharevba) after twelve
dozen circuits of the board.
7.5.7. Benin, Nigeria: Aghadaghada (Egharevba, 14). 4x4 holes, but
the four middle holes are not used; one bean in each of the remaining twelve
holes; two persons play, one being blindfolded, and has to guess to which
hole the other points. The pointer points alternately, first to a corner hole
and next to a hole in a side through this corner. After each pointing the
pointer asks kere? ‘where?’ and if the hole is guessed correctly, the bean is
removed. The correct reply for a corner is aghadagha, for a hole containing
a bean is ise, and without a bean, uo.
7.5.8. For tribe, Sudan: Aringari (R. W. Felkin, ‘Notes on the For
Tribe’ in Proc. Royal Society, Edinburgh 1885). Played by five boys on a
board set out on the ground with a number of holes arranged in a circle, and
five stores forming a cross within the circle. The boys play in turn, sowing a
stone in each hole in turn, and taking the stones in any hole when they con-
7.5.8] MANCALA. PART I 181
tain a given number of stones and transferring them to their stores. The boy
who takes most stones wins. (This may not be a mancala game.)

Or Oe@
O O
O O
C7, Os©
FIG. 82. AGHADAGHADA. BENIN

A
7-5-9. Yorubas, Nigeria: J’odu; Keta, Gold Coast: Jodorakase (K. C.
Murray). 2x6 holes and two stores, one at each end of the board; four beans
in each hole; several laps to the move; one round; play anticlockwise. A
player must leave his opponent a hole from which to play.
A move may begin from any of the player’s holes and ends when the last
bean in hand falls in an empty hole. If this hole lies in the player’s row, he
takes it and the contents of the opposite hole; if in the opponent’s row, the
move ends without capture.
7.5.10. Dahomey: Adjito (Avelot, a, citing J. Duncan, Travels in West
Africa, 1847, i. 199; and E. Foa, Le Dahomey, 1896, p. 249, from which the
description is taken). 2 4, 2X5, or 2X6 holes. The number of beans is not
given. If the last bean falls in one of the player’s holes, and the opposite hole
contains beans, these are taken.
7.5.11. Dahomey: Madji (Culin, c, 601). 2x6 holes. The board is adjito
and the beans (for which pebbles are used) adji.!

B
7.5.12. Ashanti, Gold Coast: Wari (Dr. G. T. Bennett, “The Game of
Wari’, in R. S. Rattray, Religion and Art in Ashanti, 1927, pp. 382-90). 2X6
holes and two stores, one at each end of the board; four beans in each hole;
one lap to the move; one round; play anticlockwise.
! D. Westermann, Ewefiala, 1928, p. 37, has adi, chestnut-like seeds of a vine used for playing;
J. Speith, Die Ewe-Stamme, 1906, p. 243, has adi and M. Delafosse, Manuel dahoméen, 1894, p. 357;
has aji; both of the two last works define the word as meaning mancala.
182 MANCALA. PART I [CHA7
A move can begin from any of the player’s holes, and if it is so heavily
loaded that the sowing is not completed before it reaches this hole again,
this hole is then omitted from the sowing. The player must leave his oppo-
nent a hole from which to play, and if all his opponent’s holes are empty, he
must, if possible, make a move which gives him something to play with, and
if he omits to do so, he forfeits all his beans in play to his opponent. Captures
are made when the last bean in hand falls in a hole on the opponent’s row,
making its contents two or three beans. If the hole from which the capture
is made is immediately preceded by an unbroken sequence of holes on the
same row also containing two or three beans, the contents of these holes are
also taken. When no more captures are possible, each player adds the beans
left in his own, holes to his store, and the player wins who has the larger
number of beans in his store.
The following game, played by two Yorubas in Lagos, Dec. 1945, was
recorded for me by K. C. Murray. Since all moves played by X are from
holes denoted by capitals, and all played by Y are from holes denoted by
small letters, there is no need to designate the players’ moves.
4E 2b 15F: 3c2d Ia IE te
4e 3D 2a 1D 1d 1B
4D 11a 7B : 2c 14b [Dy 2b
5c 2A Ia 7E:3c3d3e2f 1e 2E
6B 3c 3A 1b 1A 1a
Ie 2E 13f :2A3B 1D Tbe eed
6A Ia 7D :2c2d3e = Ic 1D Xwins
7b AO 2a 1K: Ic
8C:2e 11d 3h 32¢ 2a Te
26 2E: 2a Ia 3E : 2bac 1d
4E Acris, 4E Ia 1D

This is the most popular game in West Africa, and the eleven following
games are played in the same way as 7.5.12.
7-513. Gambia: Woro (Parker, 600). Both men and women play. The
stores are called woro bungo and lenko seeds are used as beans.
7.5.14. Deys, Veys, Pesseh, Gedibo, and Queah tribes, Liberia: Kpo, poo
(Culin, c, 601). The holes are called kungo and the beans kunje. The game may
also be played by three or four persons, but in this case only three beans are
placed in each hole. When three play, X owns A, B, C, D; Y owns E, F, b, a;
Z owns ¢, d, e, f. When four play, each owns half a row.
7.5.15. Ivory Coast, Agni tribe: Aware (Avelot, a); Neo tribe; Cas-
sandra: Wolo or dagboprou (Avelot, 6, 21); Kru tribe, Cavally: Wora,
oura (Avelot, a). Games undescribed; a board from the Basulo tribe is in
the Musée de l’Homme, Paris.
7.5.16. Dahomey, Nago, Yoruba, and Mina tribes: Ayo (Avelot, b).
7.5.17. Nigeria, Yoruba tribe: Ayoayo (Newberry, in Nigerian Field,
viii (1939); Shayo (K. P. Humpidge, Nigeria, 1938). The holes are called ile
7-5-17] IVIDAUIN
Gy AST Ae eyAdixaliaaT 183

and a heavily loaded hole is odu. Humpidge notes that some players take
either from holes containing two or from those containing three beans, but
not from both, settling beforehand which number shall be taken.
7-5.18. Nigeria, Okpu clan, Oturkpo, Idoma Divn., Benue Pr.: Ichwe
(K. C. Murray).
7-5-19. Nigeria, Igbirra tribe, Okeni, Kabba Pr.: Igori (K. C. Murray).
The game is named from the seeds used as beans.
7-5.19a. Nigeria, Igara tribe, Benue Pr.: Ogori (Capt. A. F. Mockler
Ferryman, 47). The only game played by this tribe.
7.5.20. Nigeria, Tiv or Munshi tribe, Benue Pr.: Dagh (K. C. Murray).
7.5.21. Gabon River, Fan tribe: Kale (Rev. A. C. Good in Culin, c, 602).
The game is named from the seeds used as beans. Usually, players do not
take the contents of an unbroken sequence of holes containing two or three
beans, but from one hole only. When the game ends, the beans left in holes
are not added to the players’ stores. Singletons cannot take. The hand when
sowing must be raised and the beans sown with some violence to prevent
cheating, and the board contains a shallow trough between the holes to catch
beans that do not fall in a hole.
7.5.22. French West Africa, Dogon tribe, Ouroli Ténné: Walu (Griaule,
b, 170). Played by men and women in the dry season. Children in Yanda-
du-bas play a four-handed variety, each child owning half a row (cf. 7.5.14).
7.5.23. French West Africa, Dogon tribe, Madougou: I (Griaule, b, 166).
2X3 to 2X7 holes, sometimes arranged irregularly on a stone board. Single-
tons cannot be moved and, if a player has only singletons, he passes his move.
Played by children.
7.5.24. French West Africa, Dogon tribe, Sanga: I pere (Griaule, b, 169).
2x5 holes; each player’s right-hand hole is his store (oho). Four beans in each
hole. Moves anticlockwise.
Each player tries to accumulate beans in his oho. The player begins by
transferring the four beans in D (d) to his oho (E, e). A move may begin from
any of his holes except his oho. If the last bean in hand falls in a hole making
its contents two or three beans, these are taken and placed in his oho. As soon
as a player has thirteen beans in his oho, he lifts and sows them, omitting
each oho, and takes the contents of any of his opponent’s holes which now
contain two or three beans. The player who takes most beans wins.
The game is played by boys and girls together during the dry season. It is
tabu to play after sunset, or a player’s mother will die unless the player
swallows a white pebble.
7.5.25. French West Africa, Dogon tribe, Madougou: Koro, ‘box’
(Griaule, b, 168). 2X 4, 2X5, or 2X6 holes; each player’s right-hand hole is
his store (hogon). Four beans in each hole; moves anticlockwise.
A variant of 7.5.24, played by boys. Each player tries to accumulate beans
in his hogon, fourteen on the 2x 4 board, fifteen on the 2x 5, and sixteen on
the 4x6 board. Moves may begin from any of the player’s holes except his
184 MANCALA. PART I [CH. 7
hogon. Singletons may not be moved. If the older player has only singletons,
he passes his move; if the younger, he lifts the beans in his hogon and sows
them as in 7.5.24, but if the opponent has only singletons, he takes them.
7.5.26. Sierra Leone, Mende tribe: Ti (F. W. H. Migeod, 281). 2x6 holes
and two stores, one at each end of the board; four beans in each hole; several
laps to the move; one round; play anticlockwise.
Only differs from 7.5.12 in three particulars; (a) the introduction of laps;
(b) each player for his first two moves only may sow them singly as in
7.5.12, OF 2, I, I or 3, 1; (c) singletons may be moved but cannot capture.
7.5.27. Sierra Leone, Mende tribe: Wari, warre (Haddon, CP.). Only
differs from 7.5.26 in that each player for his first move may sow his beans
clockwise, provided the sowing does not extend to his opponent’s row.
7.5.28. Togoland, Ewe tribe: Adi, adji, aji (Avelot, b, 11; H. Klose,
‘Musik, Tanz und Spiel in Togo’ in Globus, lxxxix (1906), 75). 2x6 holes;
four beans in each hole; one lap to the move; several rounds; play anti-
clockwise. Only differs from 7.5.12 in being played in rounds.
For the second and following rounds, the winner of the preceding round
enters all his beans in his holes, starting from A, and after completing his
own holes continuing along the opposite row from a. The last hole in which
he places beans may contain fewer than four beans. The loser then makes the
contents of this hole up to four beans and puts four beans in each of the
remaining holes in his row. These holes and the one which he completed
belong to the loser and the other holes to the winner. The game ends when
one player has lost all his beans.
7.5.29. Gabon River, Epongwe tribe: Erhérhé or Kale (Avelot, a, 267,
citing Mrs. Kate Lutterhold of Libreville, originally of Accra, Gold Coast).
2x 6 holes and two stores, one at each end of the board; four beans (tchongwa)
in each hole. One lap to the move; several rounds; play anticlockwise. Only
differs from 7.5.28 in requiring that the hole from which captures are made
contains two or four beans, and in the method of rearranging the beans for
the new rounds. The first round ends when one player has no beans in any
of his holes.
For succeeding rounds, the winner of the preceding round puts four beans
in each of his six holes, returning any surplus beans to his store. The loser
then puts four beans in as many of his holes as he can and the remainder of
his beans in the next hole. Any holes that contain no beans are ‘burnt’ and out
of play for the round.
Avelot (6, 17), writing in 1908, says that the game had been recently intro-
duced by sea, probably from the Gold Coast.
7.5.30. Cameroons, Pangwe tribe: name unrecorded (G. Tessmann, ii.
310; H.J. Braunholtz, Man, xxxi. 1931, 132). 3X7 holes without stores, or
3X § with two stores, one at each end of the board; five beans in each hole;
one lap; one round; play clockwise.
A move may begin from any of the player’s holes. If it begins from a hole
7.5.30] MANCALA. PART I 185
so heavily loaded that sowing is not completed before it reaches this hole
again, this hole is omitted from the sowing (cf. 7.5.12). If the last bean falls
in a hole making its contents two, three, or four beans, these are taken, and if
this hole is preceded by an unbroken sequence of holes in the same row, con-
taining two, three, or four beans, these are also taken.
The game is also played by the Yaunde and Bene tribes who learnt
it from tribes farther east, and more rarely by the Ntum and Fang
tribes.
7.5.31. Cameroons, M’Velle tribe: Songo, ‘pebbles’ (Major P. H. G.
Powell-Cotton, Man, xxxi. 1931, 132). 2X7 holes; five beans in each
hole; play clockwise. Only differs from 7.5.30 in that the holes from which
captures are made must contain two or four beans, or (occasionally) three
or four, or again two, three, or four beans to shorten the game.
The same game is played by the Balanga tribe on a board of 2X4 holes,
and by the Duala tribe on a board of 2x6 holes with eight beans in each
hole.
7.5.32. Nigeria, Ibo and Edo tribes: Okwe (P. A. Talbot, iii. 817). 2x 12
holes; four beans in each hole; several laps to the move; one round.
If the last bean in hand falls in a hole making its contents two or four
beans, these are taken.
7.5.33. Dahomey: Ayo (M. J. Herskovits, b, 36, citing L. Brunet and
L. Giethen, Dahomey and Dependencies, 1900, p. 333). From 2X8 to 2X 12
holes.
7.5.34. Nigeria, Tiv or Munshi tribe, Benue Pr.: Dagh (K. C. Murray)
2x 8 holes and one store.
A game on the same board is played by the Kaja tribe, Zaria Pr., Nigeria
(Avelot, a, 267, citing G. Schweinfurth, Au ceur de I’Afrique, 1875, ii. 28;
Culin, c, 602, citing Rohlfs in R. Andrée, Ethnographische Parallelen, Neue
Folge, 1889, p. 102).
7.5.35. Dahomey, Ashanti: Wari (A. W. Cardinall, 253). 2x6 holes and
two stores, one at each end of the board; four beans in each hole.
Cardinall says there are many varieties, but the commonest is played with
one lap to the move and one round. A move may begin from any of the
player’s holes, and when the last bean in hand falls in a hole making its
contents six beans, these are taken.

C
7.5.36. Gold Coast, Keta: Adi; Nigeria, Ijaw tribe: Obridjie (K. C.
Murray). 2X6 holes and two stores, one at each end of the board; four
beans in each hole; several laps to the move; one round; play anticlock-
wise.
A move may begin from any of the player’s holes, and ends when, in
playing a lap, the last bean in hand falls in an empty hole or in a hole on
either row, making its contents four beans, which are taken. Also, if at any
186 MANCALA. PART I (CH: 7
time in the course of sowing, a hole contains four beans, the player on whose
row this happens takes the four beans. The game ends when one player
can no longer move, and the player wins who has taken the larger number
of beans.
The following game, played between two Ijaw women in January 1946,
was recorded by K. C. Murray.
X 4B xGyz2e oma Xoo
5F oA ads-4BYonte.| Verb
se 3d: 4D4F 2A 2c
6D 23 AOR ID) Xen
6d 10C: 4f! 13E2 3B
Y ite X 2A Yo sh ede Y 1d:4e
7 3C 2d: 4f X 1E: 4F
7a: 4B 2F X ID and X wins
X 8C 11b 2E

7.5.37- Nigeria, Oron clan, Ibibio tribe, Oron, Calabar Pr.: Whyo (K. C.
Murray). The informant said that the name whyo came from Lagos and was
taken from the English ‘why Oh’, a slang expression referring to smart or
deceitful tricks. It seems more likely that the name is a perversion of the
Yoruba ayo, ‘mancala’.
2x6 holes neatly cut in the ground; four beans (usually palm seeds) in
each hole; several laps to the move; one round; play anticlockwise. The
game only differs from 7.5.36 in the following particulars: when the game
comes to an end, each player adds any beans left in his holes to his store, and
if another game is played, the winner of the first game now owns seven
holes, six in his own row and a on his opponent’s row, and takes fours from
his extra hole.
The following game, played by two boys, was recorded by K. C. Murray.
X 4F Va 4d 4 aD 1b Ac
sd 2D3 3a X 1B
5sC X 9A: 4B 2d Vids)
6b our Ab WE eo xaG
6B Xo, 2e Nora
Y 1b 11E, Y: 4c 3A X1D:4E
7c Yoo KaC Drawn; each player
7D 11f: 4e, X : 4F* 2D has 24 beans
8e, X: 4F MOB 4G 2

! The positions isnowo 2 I o Io I

2 2) 2 TESTO! 5
2 X, when sowing 13E, inadvertently sowed two beans in C. The position is now:
ey ee eh be i
Gert re eee EY bo)
3 Positionis9 I 0 2 2 fe) 4 Positioniso © I I 2 2
OF 3 Or Oo a9: ii, Oe Oyo Ge Sete G
7.5.38] MANCALA. PART I 187

7.5.38. Nigeria, Yoruba tribe: J’erin (R. J. Newberry, Nigerian Field,

viii (1939). 26 holes and two stores, one at each end of the board; four
beans in each hole; several laps to the move; several rounds; play anticlock-
wise.
A variant of 7.5.36. Each player for his first move lifts the beans from two
consecutive holes in his row and sows them in the usual way. When only
eight beans are left on the board, the first player lifts these and uses them to
make up the contents of any hole on the board up to four beans and takes
the two fours so made.
For the new round, the winner replaces his captures on the board, four in
each hole. Some of these holes will lie on his opponent’s row, but they now
belong to him. His opponent then fills the remaining in the same way, and
the new round is played like the first round.
7-5-39. Nigeria, Jaba tribe, Zaria Pr.: Lok (K. C. Murray) 26 holes. If
there are stores, the players do not use them, but keep their captures in a
horn. The same game as 7.5.36.
7.5.40. Northern Nigeria, Jukan-speaking peoples: ? Adzua (C. K.
Meek, b, 456). Played by women only. The same game as 7.5.36.
7.5.41. Northern Nigeria, Hausas: Darra, derra (Major A. J. N. Tre-
mearne, 57; R. Fletcher, 113; Avelot, b, 12, citing Dr. S. Passarge, Adamaua,
1895, p. 89; K. C. Murray). 2x6 holes and two stores, one at each end of
the board. Four beans in each hole. One lap to the move; one round; play
anticlockwise. Otherwise, the same game as 7.5.36.
7.5.42. Northern Nigeria, Hausas: Darra (K. C. Murray). 2X 12, 2X 16,
or 2X18 holes and two stores, one at each end of the board. Four beans in
each hole. One lap to the move; one round; play anticlockwise. Played in
the same way as 7.5.36.
7.5.43. French West Africa, Dogon tribe; Sanga, I; Madougou, Na(h)
(Griaule, b, 168). 2x5 holes and two stores, one at each end of the board.
Four beans in each hole. Several laps to the move; one round; play anticlock-
wise. Played in the same way as 7.5.36. Played by boys.
7.5.44. French West Africa, Dogon tribe, Kaca: I (Griaule, b, 168). 2x6
holes and two stores, one at each end of the board. Only differs from 7.5.43
in the board and captures. “The first player contents himself with collecting
the four beans from one hole.’ Each time a player wins a game, he occupies
one of his opponent’s holes in the next game (cf. 7.5.37).

D
7.5.45. Nigeria, Ibo tribe: Okwe (G. T. Basden, 134). A mancala game
which has reached Nigeria from Calabar and the Cross River district, and is
now played by the Ibo, Edo, Umon, Ibibio, and Owerri tribes. Okwe is the
name of the seeds used as beans.
2X10 to 2X20 holes. Boards occasionally have two stores (see Partridge,
fig. 57), but these are unnecessary as the captures are concealed in the hand
188 MANCALA. PART I [CH. 7
so that the opponent cannot see their number. Forty-three beans on each
side, arranged as follows:!
Y
ts fq po nm ck Neh ee ee decom bas
© 1020 's:0) OO 00.0 tale CS EA SS eS
Boe EOS 8) Ge eel T s Shats OO NOs OF OM Os Os OND
ASB -C DB oP GAY. [RMON OURO Ro Saee
xX

One lap to the move; one round; play anticlockwise. The game begins by
each player making an opening move; X removes Y’s beans in h, i,j, and k;
B removes X’s beans in H, I, andJ; each player keeps in his hand the beans
which he has removed, concealing their number from his opponent. The
game proper then begins. A move may begin from any of the player’s holes,
but when lifting its contents for sowing, one bean must be left in the hole
unless the hole is a singleton. If the last bean in hand falls in a hole on the
player’s row and the opposite hole contains one, three or five beans, these are
taken and added to the player’s concealed store. For any move, instead of lift-
ing beans from a hole, the player can enter some or all of the beans that are
concealed in his hand, starting the sowing from his left-hand end-hole (for
X, A; for Y, a) and proceeding along his row. If he reaches his right-hand
end-hole before he has completed this sowing, he returns to his left-hand
end-hole and continues along his row.
7.5.46. Nigeria, Ibo tribe, Aro Chuku, Calabar Pr.: Azigo (K. C. Mur-
ray). 2X 10 or 2X20 holes. The initial arrangement on the 2 x 20 board was
We
Crs? depron ml kot. Ny eerie oe ad Scopus
OF ORO O46" 0 OFO 82:02> CRE MIY 61 ote cure ncn rs
Rage Ses ee Pe Yee 6 2h 101 O) OO 0 OOO
AUD CeDeE EG oH Ilo Kol MEN-O Ph Orns ©
x

One lap to the move; one round; play anticlockwise. The players begin by
removing beans; X, ten from i, j, k, and 1; Y, nine from J, K, and L. The
position is now the same as in 7.5.45 after the opening play. The game proper
only differs from 7.5.45 in one particular: if the hole from which a capture
is made is preceded by an unbroken sequence of holes containing one, three
or five beans, these are also taken.
K. C. Murray recorded the following game which he played against an
Ibo at Aro Chuku. The score begins after the opening play.
X 4G Yue SSP sky Vise X12}
Y4g:JK X4E Y 4d X 1K Y 3h: JK
Sok nae cei X ai Yu:KL” X3H2jk
® On smaller boards the same number of holes are omitted from both ends of the board, and
the initial arrangement for the remaining holes is the same as given above.
7.5.46] MANCALA. PART I 189
Naat Vet Weick Mer aXange WAtE
Xenters14 Xul:h Xenters7 X4G Gist
Yuth:LMN Y 6e Yiseo1 Y 2d Y enters 22
X1K:i MiKcyk Xij:JKL. Nenters 3. 2K: i5..¢
7.5.47. Nigeria, Ibo tribe, Item, Bende Division (also at Abiribi, and Ona-
fia, but not at Edda): Azigo (K. C. Murray). 2x 20 holes. The initial arrange-
ment only differs from 7.5.45 in having one bean on I and one on L. X
removes nine beans from i,j, k, and 1, and Y nine beans from I, J, K, and L.
Or the stronger player removes one bean fewer, leaving the bean on | or L.
The rules only differ from 7.5.46 in two particulars: only one or three beans
can be taken from any hole, and, when a player enters beans from his store,
he must enter all of them, and if sowing reaches his right-hand end-hole
before all are sown, he has the choice of either returning to his left-hand
end-hole to sow the remaining beans in his hand, or to continue sowing
along his opponent’s row, in which case, if the hole next beyond that in
which he sowed his last bean, or an unbroken sequence of holes contain one
or three beans, he takes these. Any beans that he has so sown in his opponent’s
holes now belong to his opponent and he can lift them for sowing.
At Abiribi, only men play azigo. The women play asa but this is not a
board-game.
7.5.48. Nigeria, Akpa tribe, Benue Pr.: Okwe (K. C. Murray) 2X20
holes.
7.5.49. Nigeria, Ekoi tribe, Ikom, Ogoje Pr.: Ndim-ndum (K. C. Mur-
ray). 2X20 holes.
7.5.50. Nigeria, Ikwe tribe, Cross River: Azigo (C. Partridge, 8, 66, 210,
259 and fig. 57). Mentions of azigo as played in different parts of the Cross
River country. Two boards were seen in the Egbo House at Ntrigum.
7.5.51. Nigeria, Kukurukuland: Makwinni. A board with 2 x 10 holes is
in the Partridge collection in the Ipswich Museum.

E
7.5.52. Nigeria, Kukuba tribe, near Jos, Plateau Pr.: Iyogh (K. C. Mur-
ray). 2X8 holes. Apparently no stores since captures are held in the hand.
The board is called agumu. Four beans (iyogh) in each hole; one lap to the
move; one round; moves anticlockwise, but clockwise moves may be made
to effect a capture.
The players begin by rearranging their beans, both making the same series
of moves; X transfers all his beans in F and G to H, and three beans from E,

Ae THe; POsiHOINISIO) MON ONO MOO ONTO NOC) TONED ae lua es aS) a3 az a ONL
Oo SoG oh Se ogi oF Te ae © OG CY CC) Co
Apparently one of the beans in e had disappeared when six moves later Y only lifted six beans
from e.
ae THe POsitionsis|OmOmOn ONO MON OuOn ONO O)NEGG3 On 2arlen3 20m t
pO GD By BS PG Te ee 84 OY Oe Oy Oy OE ve
190 MANCALA, PART I [CHo7
two from D, and two from C to B; Y then makes the same transfers on his
row. The resulting position is
Va
bcos ieee dn comes
12: (0, POR) seers
42° IT 128 2 Oe Oe
JN sy, MG MIDM Meh we g(E}.. Jal
xX
When lifting beans from a hole for sowing, one bean must be left in the
hole from which they are lifted. Captures are made when the last bean in
hand is sown in a hole on either row opposite to a hole containing one or
three beans, and these beans are taken. If the next hole in the direction in
which the beans were sown, or an unbroken sequence of holes in the same
direction also contain one or three beans, these are also taken. Thus, the
game seen began by lifting eleven beans from H and sowing them clockwise
in G,F...A,h, g,...D, the hole opposite e, then contained three beans,
and C also contained three beans, so the contents of D and C were taken.
A player, when it is his turn to play, may, instead of lifting beans from
one of his holes for sowing, enter some or all of the beans which he has
already taken, beginning from his left-hand end-hole, and sowing anti-
clockwise, or clockwise if the latter leads to a capture. The game ends when
one player has no beans in his holes.

F
7.5.53. Nigeria, Benin: name unrecorded (J. V. Egharevba, 7, where it is
called ‘the concentration game’). The board (urise) has 2x 5 holes and two

FIG. 83. URISE. BENIN

stores (ogiuro) which are arranged as in Fig. 83. Ten beans (usually cowries,
but the Obas formerly used coral beads) in each hole; one lap to the move;
one round; play anticlockwise.
Moves may begin from any of the player’s holes, but one bean must be
left in the hole. Captures are made from the hole in which the move ends,
7.5.53] MANCALA. PART I I9I

and the players agree before starting to play what number of beans (two,
three, or five) this hole must contain to permit a capture (but see note 1
below). If two is chosen, the first seven captures must be of two beans; if
three is chosen, the first three captures must be of three beans; if five is
chosen, the first three captures must be of five beans. After these captures
have been made, subsequent captures must be consecutively, one bean, three
beans, five beans, &c. The game ends when one player has no beans in his
holes.
7-5.54. Nigeria, Benin: Iyagbe (Egharevba, 8). A set game on the board
of Fig. 83 with ten beans on each hole, which is played by a single person to
depict the course of an eventful war in Benin history. The holes are named
after the leading characters in that war between Edogun and Oza: A is
Edogun; B his wife, Esagho; C Olou; D Owonnokhorhopko, Olou’s wife;
E Isekherhe, Oza’s wife; a Oza; b Atakparhakpa; c Usiodamwen; d Ohom-
wannosokpa, a famous warrior; e Osuta, a parasite. In this game, in the
earlier moves at least, one, three, five, or seven beans can only be taken
from a hole though the hole may contain a larger number of beans. Only if
‘universal war’ has been proclaimed may the entire contents of a hole be
taken. Egharevba gives the score of the game, but an error after the tenth
move renders the sequel unintelligible. The first ten moves are
2d:3A 7E:3B 12a: sbscsd
AES TaN 2A 4E
4b: 1A Ire: 3A3B IoD : 3E
7B : 3bsc7d
At a later point “universal war’ is proclaimed and the game ends with the
moves 8C: all in e and A and 7B: all in C, D, E, a, b, c, and d.
From the first ten moves we see that play is anticlockwise, and that cap-
tures can be made from an unbroken sequence of holes preceding the hole in
which the last bean was sown, but the conditions which permit these captures
are not clear.!
7-5-55- Nigeria, Ibo tribe, Achalla (near Awka), Onitsha Pr.: Okwe
(K. C. Murray). 2x5 holes arranged in a circle and four stores within the
circle. Boards are marked on the ground under the verandas of houses. Ten
beans in each hole; one lap to the move; one round; play anticlockwise.
Two, three, or four persons can play.
A move may begin from any of the player’s holes, but not more than
nine beans can be lifted from any hole, and one bean at least must be left in
the hole. If the move ends in a hole, making its contents an odd number of
? This game may throw light on a feature of 7.5.53 and other games of this group, the rules
defining the number of captures of a particular value which have to be made in succession. In
7.5.53 the players agree on the number of beans that may be taken each time a capture is made.
If this means (as it would in all the preceding groups of games) that the hole from which the
capture is taken must contain that exact number of beans, this seems unworkable. Iyagbe suggests
that what the rule really means is that only a given number may be taken each time and the
hole is not necessarily emptied.
192 MANCALA. PART I [CH.7
beans, these are taken, but not more than nine beans. If a move ends in a
hole, making its contents even and the next hole has an odd number of
beans, the contents of the odd hole are taken but not more than nine beans.
If in either case the hole from which the beans are taken is followed by an
unbroken sequence of holes with an odd number of beans, these are taken.
7.5.56. Nigeria, Umuahia, Owerri Pr.: Ako okwe or Ezu ahia ako
okwe, ‘buying market’; Aro, Calabar Pr.: Okwe; Okwa, Calabar Pr.:
Akwa nsa (K. C. Murray). 2x5 holes and two stores. Ten beans in each
hole; one lap to the move; one round; moves may be made in both direc-
tions.
A move may begin from any of the player’s holes, but one bean must be
left in the hole from which the beans are lifted. Captures are made from the
hole in which the move ends; if its contents are now odd and not greater
than nine, they are taken, but the first capture must be of a single bean.
At Umuahia, the first capture must be of three beans and the second of one
bean. The game is compared with marketing and the captures are said to be
placed in the bank (nkuru).
7.5.57. Nigeria, Owerri tribe, Enugu, Onitsha Pr.: Okwe (K. C. Mur-
ray). 2X 5 holes and two stores. Ten beans in each hole; one lap to the move;
one round; play anticlockwise.
A move may begin from any of the player’s holes, but one bean must be
left in the hole, and if it contains more than ten beans, only nine can be lifted.
Captures are always made from the next hole to the one in which the move
ended. For the first seven captures only one bean can be taken, for the next
three captures, three beans must be taken, and for the next five captures only
one bean is taken. The first seven captures can be made from holes on either
row, but subsequent captures can only be made from holes on the opponent’s
row. (There is some rule about playing from singletons which was not
obtained.)
7.5.58. Nigeria, Bende tribe, Enugu, Onitsha Pr.: Okwe (K. C. Murray).
Only differs from 7.5.37 in allowing moves to be made in either direction.
7.5.59. Nigeria, Ibo tribe, Item, Bende Division: Ekwe (K. C. Murray).
2x5 holes. Ten beans in each hole. Three persons play.
7.5.60. Nigeria, Owo, Ondo Pr. When K. C. Murray inspected the
stores of the Afin Oluwo, he found a circular board with twenty-four holes
round the circumference and four stores in the middle, but no one remem-
bered how the game was played.
G
7-5.61. Nigeria, Umon, Aro Division, Calabar Pr.: Gifta (K. C. Murray).
2X4 holes and two stores, one at each end of the board. Each player owns
the two holes in each row and the store to his left side. Six beans in each hole;
one lap to the move; several rounds; moves can be made in either direction,
and go round the board in the usual way.
7.5.61] MANCALA. PART I 193
A move may begin from any of the player’s holes. When a player has only
a few beans left in his holes, he may refrain from moving when it is his turn
to do so. If a move ends in a hole making its contents two or four beans, he
takes these, and if the adjacent hole on either side also contains two or four
beans, he takes them also. If a move ends in a hole making its contents other
than two or four beans, but the adjacent hole on either side contains two or
four beans, he takes from one of these holes, but not from both. When each
player is reduced to one singleton, the round ends and each player takes the
singleton in his own row.
Each player then arranges his beans in his own holes, putting six in as many
of his holes as he can, and returning any surplus beans to his store. The round
is played in the same way, and rounds are played until one player has lost all
his beans.
K. C. Murray recorded the first round of a game played by two women in
Umon. I use the following notation for the holes:
Y
D2 Ce Ube 3
Av SB} cr
x

and clockwise moves are marked with an asterisk.

X 6B YE THEY SAD X 3C Y misses turn
Y*7b X 11A: 4d2a Ye3becc X oB : 2d2a
X 8D Yesbs32D2A X*5A : 2d2a? Y 4c: 2b2C
Y od:2b X 14c:2d2a Y misses turn X 2A
X 2D:4B .Y¥*15¢ 3 2d2a' X 6D: 2b Y misses turn
X 2B: 2c2d
X has now twenty-nine beans (including one in D) and Y has nineteen
(including one in a). The beans were now rearranged for the next round thus:

nN No lo

and X retains five beans in her store and Y one in her store. The remainder
of the game was not recorded.
7.5.62. North Cameroons, Nsungli tribe: Mbangbi (J. Sieber, in Afrika,
xi. 1938, 208-20).
2x5 holes and two stores, one at each end of the board. Eight beans in
each hole. Two or four persons play.
If in the course of sowing the last bean in hand is sown in a hole making
its contents two or four beans, then these and any immediately preceding
unbroken sequence of holes also containing two or four beans are taken.
’ The positionis4 3 3 0 2 The positionis6 Ir I oO
Ao 7 ean o0 8 0 Oo
5856 oO
194 MANCALA PART I [CH.7

H
7.5.63. Sudan, northern tribes: Mangala or um el-bagara, ‘cow game’
(R. Davies, SNR., viii. 140). 2X5 holes. Five beans in each hole. Moves
from A, B, a, b are made clockwise, from D, E, d, e anticlockwise, and from
C, c can be made in either direction.
Singletons cannot be moved. If the last bean of a move falls in a hole
making its contents two or four beans, these are taken, and if this hole is
preceded by an unbroken sequence of holes also containing two or four
beans, these are also taken.
7.5.64. Sudan, northern tribes: Mangala or Um el tuweisat, ‘little goat
game’ (R. Davies, SNR., viii. 141). 2X3 holes. Three beans in each hole.
Played by children. Moves from A, a are made clockwise, from C, ¢ anti-
clockwise, and from B, b in either direction. The game is played in the same
way as 7.5.63.
I
7-5.65. Nigeria, Ibo tribe, Achalla (near Awka), Onitsha Pr.: Okwe (K.C.
Murray). In addition to 7.5.55, there is a local variety of mancala, played on
a board of 4X4 holes with, at times, an odd hole at one end. The observer
did not see the game played, but was told that the odd hole and two in the
middle were simply stores.
7.5.66. Sudan, Baggara tribes: Kara (R. Davies, SNR., viii. 142). 4X4
holes, the four in the middle being stores. Played by four persons, each own-
ing a corner hole and the two holes on either side of the corner. The same
number of beans in each hole.
If a move ends in the player’s right-hand end-hole or in an opponent’s
holes, making its contents two, four, or six beans, these are taken.
7.5.67. Nubia: Mangala, moungala (Avelot, a, citing Schweinfurt, Au
Coeur de I’Afrique, 1875, ii. 28, and Linguistische Ergebnisse einer Reise nach
Central Afrika, 1873). 3X8 holes. Small pebbles used as beans.

6. EAST AFRICA: ABYSSINIA, SOMALILAND, SUDAN, KENYA, AND

UGANDA

The mancala II games played in East Africa, north of the Equator, show
marked differences from those played in West Africa, chiefly in the syste-
matic use of both directions of play, and in the treatment of holes whose
contents are made up to four beans in the course of the games. For the latter,
we may compare the analogous treatment of these holes in the Ceylon game,
puhulmutu, 7.3.1.
7.6.1. Abyssinia: Gabata (M. Cohen, Journal Asiatique, 10 Ser., xvili
(1911), 491-7); Guvoota (W. C. Plowden, 273, 276); Gabatta (J. T. Bent,
a, 72); Guebeta (Avelot, b, 12).
There is a reference to mancal as an obsolete Abyssinian game in Father
Francisco Alvarez, Verdadeira Informagam do Preste Joas das Indias, Lisbon
7.6.1] MANCALA. PART I 195
1540, which gives an account of the Portuguese embassy to Abyssinia in
1520-5 (Culin, c, 601). Probably Alvarez refers to mancala II which seems
to have been replaced by mancala III, for Bent found the latter game widely
played. Cohen (p. 391), however, says that the only mancala game now
played is mancala II, though those who still possess boards for mancala III
use them for mancala II by disregarding the middle row of holes. Several
varieties of gabata are played, but more often by adults than by children.
7.6.2. Abyssinia: Uugg (Cohen, 491). 2X5, 2X6, or 2X7 holes and two
stores, one at each end of the board. Four beans in each hole. Several laps to
the move; several rounds; play always anticlockwise.
A move may start from any loaded hole on the player’s row, provided it
is not a uugg. If the last bean in hand falls in an empty hole, the move ends; if
in a loaded hole on either row, except one on the opponent’s row which now
contains four beans, the contents of this hole are lifted for a new lap; if in a
hole on the opponent’s row making its contents four beans, this hole becomes
a uugg and its contents are left in the hole to accumulate by future sowings,
until at the end of the round when the player who made it a uugg adds its
contents to his store. If a uugg is preceded by a singleton, the player who owns
the row on which the uugg stands, may begin by lifting the singleton and
sowing it in the uugg and then take two beans from the uugg and add them
to his store. When no further moves are possible, the round ends.
For the next round, the player with most beans arranges them on the
board, putting four beans in each of his own holes and continuing along his
opponent’s row until all are entered. His front hole, now on the opponent’s
row, may contain fewer than four beans, and his opponent makes its contents
up to four and puts four beans in the remaining holes on his row that are
empty. If he has to use one bean to complete the first player’s front hole, the
hole belongs to the first player, if two or three, the hole belongs to him (cf.
7.5.28 and 7.6.8). The new round is played in the same way as the first
round.
7.6.3. Abyssinia: Sullusie (Cohen, 497). A variety of 7.6.2. in which
holes containing three beans are preserved, but Cohen did not know how.
7.6.4. Abyssinia: Qancebie (Cohen, 497). A variety of 7.6.2. in which
there is prescribed opening play. The beans are first arranged, four in each
hole, and then each player takes one bean from his alternate holes (A, C, E,
&c.) and puts each in the next hole, so that the players have from left to
right holes with three and holes with five beans alternately.
7.6.5. Somaliland: Togatog or Saddigqiya (R. F. Burton, First Footprints
in East Africa, 1856). 2X5 holes. Four beans in each hole. “When in the
course of play four men meet in the same hole, one of the adversary’s is
removed.’ Played by children.
7.6.6. Somaliland: Leyla-gobale (G. Marin, a, 506). 2X6, 2X8, or
2x12 holes. Four beans in each hole; several laps to the move; one round;
play clockwise only.
196 MANCALA. PART I [CH. 7
The first player begins by lifting his beans in F and sowing them clockwise,
the whole move consisting of the laps 4F, 5B, 5c, 6D, 6d, the last bean in
hand falling in D and making it a singleton. He now removes the contents
of D and c and puts them in his store and the opponent plays. He may begin
from any of his holes, and this is the rule for beginning any move, except in
so far as it is modified by the rules for captures. A lap ends when the last
bean in hand is sown in an empty hole. If this hole is on the opponent’s row,
the move ends. If it is on the player’s own row and the opposite hole is
empty, the move ends. If the opposite hole contains one, two, or four beans,
these and the last bean to be sown are taken; but if it contains three beans,
one bean is transferred from this hole to the hole in which the last bean in
hand was sown, making the contents of both holes two beans. Each of these
holes becomes an “ur which belongs to the player who made them ‘urs. No
bean can be lifted from an ‘ur to initiate a lap or move, but beans are sown in
them in the ordinary course of the game and they accumulate for the benefit
of the player who owns the ‘ur. When one player has no beans in his other
holes, the game ends. Each player then collects the contents of his “urs and
other holes on his row, and the player who has taken the larger number of
beans wins.
7.6.7. Somaliland, Darod tribe: Bosh (G. Marin, a, 506). 2X5 holes.
Four beans in each hole; several laps to the move; several rounds; play anti-
clockwise only.
A lap begins, not from the hole in which the last bean of the preceding
sowing was sown, but from the hole next in front of this. If this hole is
empty, the move ends, but the player takes the contents next beyond it.
When one player is left without beans in any of his holes, the round ends
and the opponent takes any beans left on the board.
For the new round, each player proceeds to fill as many as he can of his
own holes by placing four beans in each, and returns any surplus beans to
his store. Any hole which neither player has so filled is out of play for the
round.
7.6.8. Sudan: Um el banat, ‘game of daughters’ (R. Davies, “Some
Arab Games and Riddles’, in SNR., viii. 143). 2X6 holes. Four beans
in each hole; several laps to the move; several rounds; play anticlock-
wise.
A lap ends when the last bean in hand falls in an empty hole or makes its
contents four beans when the player is said to have begotten a daughter, in
which case he marks the hole in some way to show his ownership of it. A
lap cannot begin from a daughter-hole, but beans are sown in it in the usual
way and these accumulate for him. If the last bean in hand is sown in a
daughter-hole belonging to the sower, his move ends, but if in one of his
opponent’s daughter-holes, he removes two beans from it and adds them to
his store, and then starts a new lap from any of his holes except daughter-
holes. Thus in the. position
7.6.8] MANCALA. PART I 197

where a and f are X’s daughter-holes and it is Y’s turn to play, Y plays
2d : 2f and then re: 2f, reducing X’s daughter-hole f to a singleton.
When the only beans left on the board are in daughter-holes, each player
collects the contents of his daughter-holes and adds them to his store. The
player with the larger number of beans then proceeds to rearrange his beans,
starting from his left-hand end-hole and putting four beans in each hole and
continuing along his opponent’s row. His front hole may contain three, two,
or one bean. The opponent then makes up the contents of this hole to four
beans; if this uses one bean, the hole belongs to the first player, and if two
or three it belongs to the opponent (cf. 7.5.28 and 7.6.2). The opponent then
puts four beans in his remaining holes, and the round is played in the same
way as the first round.
7.6.9. Eastern Uganda, Ja-luo tribes of Kavirando: Name unrecorded
(C. W. Hobley, 6, 34). 2 x 8 holes. Three beans in each hole. ‘A kind of bao.’
7.6.10. Kenya, Kikuyu tribe: Kiuthi, giuthi (L. B. S. Leakey, 165). 2x 5
or any number up to 2X10 holes and two stores, one at each end of the
board. Six beans in each hole; several laps to the move; several rounds; play
alternately clockwise and anticlockwise.
A move may begin from any of the player’s holes except a singleton, and
in either direction. If the sowing fails to reach a hole on the opponent’s row,
the player makes a second move from one of his loaded holes and in the
opposite direction, and continues to do so until he has sown in holes on his
opponent’s row. If now the last bean in hand is sown in a loaded hole, its
contents are lifted and sown in the opposite direction to that of the preceding
lap and further laps are played until the last bean in hand falls in an empty
hole. If this hole lies on the opponent’s row, the move ends; if on his own
row after sowing beans in all his opponent’s holes and the opposite hole con-
tains beans, he takes its contents and the bean he has just sown in his own
hole. If the last bean in hand is sown in an empty hole on his own row and
the next hole on either side is also empty and the opposite hole contains
beans, these are also taken. The round ends when both players have only
singletons left. Each player takes his own singletons and puts them in his
store, and if both players now have an equal number of beans, the game is
drawn. If the numbers are unequal, the beans are rearranged for a new round.
The player with the fewer beans arranges his beans in his holes in any way
he likes, but putting at least one bean in each hole, and his opponent copies
this arrangement in his holes, returning any surplus to his store. If the loser
has less than half his original number of beans, he shortens the board by one
hole at one end of his row, if less than a quarter he may reduce the board to
23 holes, and if less than one eighth to 2X2 holes. The round is then
198 MANCALA. PART I [CH. 7
played on the reduced board in the same way as the first round, and if the
loser regains sufficient beans to make their number more than half his original
number, the next round must be played on the complete board.'
Dr. Leakey gives the opening moves of a game on the 2x6 board:

X*6F X*11D:17Ebe X 2F *2B

af warts *ob:s5Fb 6f
*7a 736 Y AG *2F
Y 6e 2b *r4A 4D
*I0E:3fF *11d:27BCb 3e *2b
2F : 6Eb
ending in the positiono 0 2 I oO 4
TSG ey oe ce)
X having twenty-two in his store and Y thirty-six.
If a player is caught cheating, his opponent takes all the beans on the
board. In some places, his opponent may insist on the move being replayed.
7.6.11. Kenya, Kikuyu tribe: Kiuthi (L. S. B. Leakey, CP. An earlier
account than 7.6.10). 2X 8 holes. Four beans in each hole. The only difference
in this account is that when making a capture, if the hole from which it is
made is followed or preceded by an unbroken sequence of empty holes, the
beans in the holes opposite these holes are taken.
7.6.12. Kenya, Akamba tribe: Kiusu, kiuthi (G. Lindblom, 427). 2x 5
to 2 10 holes. Four beans in each hole. “The object of the game is to end a
move in an empty hole, by so doing the player takes the pieces in the hole
opposite.’
7.6.13. Kenya, Masai tribe (Sakirari only): Kiuthi (L. B. S. Leakey, CP.).
28 holes and two stores, one at each end of the board.
Four beans in each hole except the two end-holes of each player’s row.
Several laps to the move; several rounds; until the player makes a capture,
play is anticlockwise except to make a ram; after his first capture it is alter-
nately anticlockwise and clockwise. Singletons cannot be moved.
A move may begin from any of the player’s loaded holes, but must extend
to his opponent’s row. If in the course of sowing, the last bean in hand falls
in a hole on the opponent’s row, making its contents four beans, the hole
becomes a ram. Beans are sown in a ram in due course, but beans in a ram
cannot be lifted for a lap. If the last bean in hand of the player who made the
ram falls in the ram, he takes its contents, but at the end of the round the ram
belongs to the player on whose row it lies and he adds its contents to those
in his store and uses them when he rearranges his beans for a new round.
The rules of capture and for new rounds are the same as in 7.6.9.

In some places when playing later rounds, the player with the fewer beans is not obliged to
arrange all his beans, but may retain some in his store, and when the board is reduced to 2x 2
holes, he may insist on his opponent copying his arrangement, not vertically as is the usual rule,
but diametrally opposite.
7.6.14] MAN@AITAG] PART 199

7-6.14. Kenya, Masai tribe: En dodoi, en geishei (C. Merker, 36 and

Fig. 8; Avelot, b, 13). 2X6 to 2X10 holes (but Merker’s figure shows 2x 5
holes). Four beans (seid) in each hole; several laps to the move; one round;
play anticlockwise only. Played by men only.
A move may begin from any of the player’s holes. If the last bean in hand
falls in an empty hole on the player’s row and the opposite hole contains
beans, the contents of both holes are taken.
The Asa (Wanderobi) do not play en dodoi because it leads to neglect of
the hunt and lack of food. The Masai say that en dodoi was invented by
Sindillo, the son of the first man, Maitoumbe, and was originally called geshe.
The game is played for an agreed number of points, each win counting
one. The score is kept by ivory pegs (alai), 15 to 20 cm. in length.
7.6.15. Kenya, Masai tribe: Guehe (Merker, 36). 2x 40 to 2x 50 holes,
marked in the ground. Four beans in each hole. Played by two sides, each of
eight warriors, all of one side playing in succession. The same game as 7.6.13.
7.6.16. Kenya, Embu tribe: Uthi (Orde Brown, 126). 2 8 holes and two
stores, one at each end of the board. Five beans in each hole, several laps to
the move; one round; though if there are not enough beans, the front hole
may contain fewer or be left empty; several laps to the move; one round;
play alternately anticlockwise and clockwise.
The game begins by each player lifting the beans from one of his holes
and putting three beans in the next hole in either direction and two in the
hole beyond it. The second and following moves may begin from any of
the player’s holes, the player sowing them one by one in the succeeding
holes, playing in either direction. The next lap is played in the opposite
direction. If sowing extends to the opposite row and returns to the player’s
own row, and ends there in an empty hole, the player takes the bean from
this hole and the contents of the opposite hole. The game ends when one
player has lost all his beans.
7.6.17. Uganda, Luganda: Mweso; Lunyoro (east of Lake Rudolf): Oku-
sora; Teso (east of Lake Kiogo): Akileyisit (A. L. Kitching, 227). Two
rows of holes. ‘Under certain conditions the player can take all the men in
the hole opposite to the one in which he concludes, and distribute them on
his own side of the board.’
7.6.18. Uganda, Karungi: name unrecorded (Sir F. Treves, 170). A photo-
graph showing two boys playing on a board of 2X8 holes with stores at
either end of the board.
7.6.19. Kenya: name unrecorded (Elspeth Huxley, The Sorcerer's Appren-
tice, 1948). Three brief references: p. 16 (Mambrui, Kenya). ‘In their small
verandas men sit all day, some at sewing machines, some playing the inter-
minable game of bao, a board between them dented with twin rows of little
holes into which they drop the beans’: plate opposite p. 94 shows two
Masai warriors playing ‘bao’ on a board of 2x10 holes without stores;
p. 161 (Nzega, 150 miles south of Mwanza and Lake Victoria): “Two young
200 MANCALA. PART I [CH. 7
men were playing bao (that universal and incredibly complicated kind of
African backgammon) on a bench outside the thatched white-plastered hall.’
(Here bao, really the name of a mancala IV game, is loosely used for
mancala II games.)
7.6.20. Kenya, Samburu tribe (south-east of Lake Rudolf): Intotoi (Miss
Grindley). 2x 12 holes. The initial arrangement of the beans is:
353. 0. 3 5 ARO 3 re Os or one”
OF Std OR Fahad > Cire oe meee Oe Seen
A move may comprise several laps but only if the last bean in hand falls in a
loaded hole on the opponent’s row. Play is anticlockwise.
Opening play. X lifts two beans from K and sows them in L and a. Y then
lifts three beans, either from k or from 1, and sows them, starting new laps
until the last bean in hand falls in an empty hole. This will necessarily happen
on X’s row. This concludes the opening play.
In the subsequent play, a player may begin his move from any hole on his
own row, but only if the sowing ends in an empty hole on his own row or
extends to his opponent’s row. If it ends in a loaded hole on the opponent’s
row, its contents are lifted for a new lap, and further laps are played until the
last bean in hand falls in an empty hole; if this hole lies on the player’s own
row and the opposite hole contains beans these are taken.
7. GAMES PLAYED IN CENTRAL AFRICA
I know of no descriptions of the mancala games which are played in the
Bahr el-Ghazal and the neighbouring districts to the north of the Belgian
Congo. G. Schweinfurth, Artes Africanae, 1873, ii. 28,! says all the tribes of
the Bahr el-Ghazal, except the Mangbattus, play mancala II, and gives a
drawing of the board used by the Niam-niams. M. C. Overbergh, Mangbetu,
1909, p. 437, says that the Mangbattus now play mancala II, and Avelot (b, 20)
says that the game is travelling westwards along the valley of the Welle,
and gives the following list of the names by which the game is known:
Niam-niam eee ; ; i . Abanga
Bongo . : : : » Lond
Shilluk (Diour). oh yes el edat et r Ee!
Mangbattu (Mombutto) . , ; . Mémbille
A-Madi . ; : ; : : . Ambiddo
A-Bangba ‘ : : : : ene
A-Ndakko ; : : : : . Mbéré
A-Barambo_. é . 2 : pe Gcue

8. GAMES PLAYED IN THE CONGO BASIN AND ANGOLA

Mancala II games appear to be played throughout these regions, but I

have been able to find very little about them, and few of these mentions make
it clear whether the games are mancala II or mancala IV.
? Cited by Avelot (4, 267) and Culin (¢, 602). See also Avelot (b, 20).
7.8.1] MANCALA. PART I 201
7.8.1. Congo Basin: Bao (Sir H. Johnston, a, 121). Grenfell saw bao at
Lobenja. It was played with beans or pebbles on a board on a thick piece of
wood, scooped out into a number of shallow holes. He does not say how
many rows were used, and, if he is giving the native game, which is not
certain, it may be the mancala IV game played in Zanzibar.
7.8.2. Congo Basin, Bangi and Yanzi tribes: Bonzo (Sir H. Johnston, a,
709). “A kind of draught-board in which two rows of parallel holes are
scooped, the counters being pins or kauries.’
7.8.3. West Congo, Konjo tribe: name unrecorded (Sir H. Johnston, a,
798). Dr. Holman Bentley gives a list of West Congo games in his Dictionary
of the Konjo language.
7.8.4. Congo Basin, Urundi (north end of Lake Tanganyika): name
unrecorded (V. d. Bergh, Dict. Frangais-Urundi, 319).
7.8.5. Congo Basin, Boloki (Bangala) tribe: Peki (J. H. Weeks, 153). The
vaguest of descriptions, but Weeks says that few Europeans understand the
game.
7.8.6. Congo Basin, Urua: Kisolo (Avelot, a, 267, citing V. L. Cameron,
A travers l’Afrique, 1878, App. 542). The Africans of Urua play kisolo (for
this name cf. 8.2.7, 8, 9, 18) with grains called masoko.
7.8.7. Angola, Kimbono; Nano, Lobale, Lounda, Boumbe, and Ovampo
tribes: Chela (Avelot, a, 267, citing P. Bouche, Sept Ans en Afrique occiden-
tale; Avelot, 6, citing L. Magyar, Reisen in Sud-Afrique, 1849-1857 aus dem
Ungarischen von J. Hunfalvy, 1859, i. 314-15). ‘Chela is a rectangular prism of
wood, pierced with forty small round holes, disposed at equal distances on
two rows, each hole being filled with round objects, usually the seeds of
forest fruits. Each player takes in his hand three of these seeds and throws
them like dice on the pad in the middle of the board whence they roll into
the holes on one or other row, causing loss or gain. If the seeds played by one
player fall in any of his opponent’s holes, they belong to the latter.’ Avelot
was doubtful whether this game was mancala or not, but it may easily be a
traveller’s generalization from a casual sight of mancala.
7.8.8. Angola: Mbau (Culin, c, 602). 2x6 holes and two stores, one at
each end of the board. A board from Elmina, collected by Héli Chatelain,
who lived for some time in Angola, is now in the U.S. Nat. Museum,
Washington, and is figured by Culin. It has a strong resemblance to the
West African boards. Chatelain told Culin that boards were often cut on
rocks near stations where porters halted.
9g. AMERICA. GAMES PLAYED BY NEGROES IN THE WEST INDIES,
GUIANA, AND THE SOUTHERN STATES OF U.S.A.
The Slave Trade between the Guinea Coast of Africa and the West
Indies and the neighbouring parts of North and South America introduced
thousands of African slaves into America who brought with them their social
habits and games, and their descendants still play mancala II games widely in
202 MANCALA. PART I [CH. 7
these areas. M. J. Herskovits has published a valuable study of the existing
mancala games of these peoples in his “Wari in the New World’ in JRAI.
Ixii (1932), 23-37.
In Dutch Guiana, the game is generally associated with the houses of
mourning in which the dead are placed during the ten days’ period preceding
burial; boards are part of the furniture of these houses, and the game is
played during vigil by the mourners to amuse the spirit whose body awaits
burial. The making of mancala boards is no small thing and is reserved for
men who have lost their wives or attained old age. Two types of board
exist, one with a plane surface, the other with a curved surface which has
some resemblance to a boat, both having 25 or 2X6 holes with stores at
both ends of the board. The houses of mourning usually contain boards of
both shapes, so that, if a person who preferred one form of board dies, the
mourners may play for a time with the other, and so reduce the chances of
the dead man’s spirit coming to play with them.
In the West Indies, the game is also associated with wakes for the dead,
and there is a general and strong objection to playing mancala after dusk for
fear of attracting the attention of spirits.
The game is known as wari in British Guiana, Trinidad, Barbados, St.
Lucia (also wa-wee), Grenada, Domenica, Antigua, St. Kitts, Haiti, San
Domingo, Martinique, and Louisiana; awari in the coastal regions of Dutch
Guiana, or adji-pre and adji-boto by the Bush negroes of Dutch Guiana (cf.
adji, aji in the Ewe language, see 7.5.11).
The games are essentially games for men, though women sometimes play,
but only with other women.
7.9.1. Barbados: French wari (Herskovits, b, 32). 2x6 holes and two
stores. Four beans in each hole; several laps to the move; one round; play
anticlockwise.
A move may begin from any of the player’s holes. If the sowing ends
in a hole containing beans, its contents are lifted for a new lap; if in an
empty hole and the opposite hole contains beans, these are taken (cf. 7.5.9
and 10).
7.9.2. Barbados and other British West Indies: English wari (Herskovits,
b, 30). 2X6 holes and two stores, one at each end of the board. Four beans in
each hole. This is essentially the same as the African wari (7.5.12), but there
are some local variations. Thus, in St. Lucia, each player may for his opening
move lift the beans from any of his holes except the right-hand end-hole, and
place them in the next hole on the right, and when sowing the contents of a
heavily loaded hole and reaching the same hole again, it is not omitted from
the sowing.
7-9.3. Dutch Guiana, coast near Surinam: Awari (Herskovits, a, 90,
citing Van Panhuys, Encycl. van Nederlandsch West Indie, 163). Adji-pre, Langa-
holo (Herskovits, b, 28). 2x6 holes and two stores, one at each end of the
board. Four beans in each hole; play anticlockwise. This is essentially the
7.9.3] MANCALA. PART I 203

same as 7.5.12, except that a player who is unable to move when it is his
turn to do so, passes his move until he can move.
7-9.4. Dutch Guiana, Saramacca tribes of Bush negroes: Adjiboto, ‘man-
cala boat’ (Herskovits, a, 90; b, 25). 2x5 holes and two stores, one at each
end of the board. Ten beans in each hole; one lap to the move; several
rounds (though the players may agree to play one round only); play anti-
clockwise (cf. 7.5.53-57). Singletons may not be moved.
Opening play. Each player, playing alternately, makes five moves of a
single lap, starting each move from a different hole in his row. During these
opening moves, no captures can be made.
After these moves have been made, a move may begin from any loaded
hole on the player’s row. If the lap ends in a hole on either row preceding a
hole which contains one, three, or five beans, or an unbroken sequence of
similarly loaded holes, these are taken and the move ends. If at any time one
player is unable to move in his turn, his opponent continues moving until
he is able to move again. When neither player can move, the round ends,
and each player collects the singletons in his holes.
The beans are then arranged for a new round. The winner of the preceding
round puts ten beans in each of his holes, and the loser puts ten beans in as
many of his holes as he can, starting from his left-hand hole. The winner then
returns any complete set of ten beans to his store; then the two players
exchange any beans that are left over and add them to their stores. Any empty
holes are akuna. Beans are sown in akuna holes in the course of play, and if a
move ends in an akuna hole by either player, he takes its contents, but no
move may be begun from any akuna hole. The new round is played in the
same way as the first round, including the opening play from each of the
loaded holes.
7-9.5. Dutch Guiana, Djuka tribe of Bush negroes: Adji, adji-pre,
lontu-holo (Herskovits, b, 25). 2x6 holes and two stores, one at each end
of the board. Since each player owns one half of each row (cf. 7.5.60), I use
the following notation:

F D c
A C d p
hr

tt
wo
~~
x Kook
Four beans in each hole and each player’s beans circulate anticlockwise
round his holes. The aim of each player is to get rid of his beans by trans-
ferring them to his opponent’s holes. Beans can only be transferred to loaded
holes. So far as I know, no similar mancala game is played in the Old World.
The game begins by a series of opening moves. There are many varieties,
but Herskovits only learnt four. Apparently, the first player selects the form
of opening and his opponent repeats it on his half of the board.
1. X transfers all the beans in D, E, and F to B. He then sows 4C and next
12B, ending in F. He then picks up the four beans in each of D, E, and F
204 MANCALA, PART I [CH.7
and holds them in hand while Y repeats these opening moves on his half of
the board. X’s position is now (A-F), 7, 2, 3, 0, 0, 0. X then enters the beans
he holds in hand, four in each of a, b, and c. Y does the same in A, B, and C.
2. X transfers 2D and 2F to E and 2A and 2C to B; sows 8B and the
position is (A-F) 3, 1, 4, 4, 9, 3. He picks up 4C and 4D and holds them in
his hand while Y repeats the moves. He then puts four of the beans in hand
in each of two of Y’s loaded holes. Y repeats this.
3. X transfers 2A, 2B, 2D, and 2E to C and 4F to A, B, D, and E (one
bean in each), making three beans in each of these holes; next 3E to A, B, D,
and the position is (A-F), 4, 4, 12, 4, 0, 0. He then picks up the contents of
A, B, and D, holding the twelve beans in hand while Y makes the same
moves. X then transfers 2C to one of his empty holes. Y does the same and
first X and then Y puts the twelve beans in hand in one of the opponent’s
loaded holes.
4. X transfers 3D and 1A to B; 3C and 1D to E; 1D to A and IC to F,
and picks up 4A and 4D and holds them in hand while Y makes the same
moves. X then puts four beans in hand in two of Y’s loaded holes, and Y
does the same.
When the opening play is finished, the players play in turn, lifting and
sowing the beans from one of their holes and omitting this hole during this
sowing. If the sowing ends in a hole making its contents four beans, these
and those in an unbroken sequence of fours are placed in any of his oppo-
nent’s loaded holes, and if his opponent has only singletons, he transfers one
singleton to another, making a hole with two beans, and transfers the beans
to this hole.
 8
NUAUN GALS PAA
Re eer T
IN this chapter I deal with those forms of mancala which are played on boards
with three and four rows of holes. They are now only played in east and
southern Africa, though there are grounds for believing that both boards
were invented in Arabia.
MANCALA III
The earliest references to mancala with three rows of holes that I have
found occur in Arabic works of jurisprudence in which the legal status of
games is examined. The greater part of these works relates to chess and nard,
but occasional references are made to other board-games, qirg, ‘merels’ and
a game called arb‘ata ‘ashara, ‘fourteen’, e.g. by b. Taimiya (died 1328) in
his Qaida fi'l-li*b bi’sh-shatranj (MS. Constantinople, Omoumi 1001, p. 7),
without any explanation what the game of ‘fourteen’ was. A later writer,
al-Qabiini (died 1464), however, when quoting from ar-Rafi‘'i (died 1226),
in his An-nasikha li’l hurr wal-‘abd bijtinab ash-shatranj wan-nard (MS. Berlin,
$498) thought it necessary to explain ‘fourteen’ and added the note (f. 61a)
‘Take a piece of wood and dig in it three rows of holes in which are placed
small pebbles with which it is played’, which shows that ‘fourteen’ was a
mancala game.
Both of these writers base their discussion on the legal status of this game
on a tradition that abi Hanifa (died 767) had once remarked that chess, nard,
and fourteen were clearly haram, i.e. forbidden, an opinion that was subse-
quently modified by the less drastic schools of lawyers, who regarded these
games as only makrih, i.e. disapproved but not under penalty, so that
players of these games incurred no legal disabilities.
Mancala III is apparently no longer played inArabia, and has never been
played in other parts of Asia. Its diffusion took a westward direction, first to
Abyssinia and then to neighbouring African regions, its farthest penetration
seems to have been to the Bushongo tribe in the Congo basin, where a board
of 3X3 holes with stores at each end of the board appears on the statue of
Bolougongo, who is said to have introduced the game in the seventeenth
century. It was still played in Abyssinia in 1873, when J. T. Bent saw it and
secured two boards which are now in the British Museum, but it had dropped
out of use by 1911 (Cohen. ‘Jeux Abyssin’ in Journal Asiatique, oth ser.,
Xviii). So far as I can learn, it only survives in parts of Kenya and Uganda,
but no observer has given any details of how it is played.
! The analogy of mancala IV games suggests that each player owned the outer row on his side
of the board and that men circulated anticlockwise round this and the middle row, captures
being made from the middle row only. Trial shows that this gives a playable game.
206 MANCALA. PART II [CH. 8
8.1.1. Arabia: Arb‘ata ‘ashara, ‘fourteen’. Board with three rows of
holes.
8.1.2. Abyssinia: Gabata (J. T. Bent, Sacred City of the Ethiopians, 1893,
72; Avelot, b, 12; Cohen, 491). 3X6 holes (toukoul, ‘huts’) and two stores,
one at each end of the board. Three beans (chachtma) in each hole. “The game
is played by a series of passing, which seemed to me very intricate and which
we could not learn’ (Bent). This is probably one of the games, gittee and

FIG. 84. BOARD FOR GABATA (Abyssinia). From Bent,

Sacred City of the Ethiopians

guvoota which W. C. Plowden (Travels in Abyssinia, 1868, pp. 273, 276)

played in Abyssinia in 1843-8.
8.1.3. Somaliland: name unrecorded (Avelot, b, citing C. B. Michel,
Vers Fachoda, 492). ‘A game similar to gabata, but more rudimentary.’
8.1.4. Eastern Uganda, Nandiek or Nadi and Lumbwa tribes: Komari
or Bau (C. W. Hobley, 1902, 6, 42). 3 x6 holes. “The children play a sort of
Komari or Bau with Stones moved along a number of small depressions in the
ground; there are usually three rows of six such depressions in a row.’
8.1.5. Congo Basin, Bushongo tribe: Lela (E. Torday and T. A. Joyce,
97). The mancala game now played by the Bushongo tribe has four rows of
holes, but mancala III is shown on the statue of Shamba Bolougongo, as
stated above. It is of course possible that the actual board may have con-
tained more than the three holes in a row which is shown on the statue.!
MANCALA IV
The earliest mention of a mancala board with four rows of holes that I
have found occurs in Hyde (p. 232): ‘among the Hanzoannites is a game
which in their language is called morahha (as Abdallah Shah, their king’s son
and envoy to the king of England told me) which is not unlike the preceding
game (mancala) and is also played with pebbles. But in its plan, there are
four rows of holes which lengthwise are eight in number; sixteen on each
' Parker (p. 591) says that he had seen some Madagascar boards of 3 x 8 or 3 x9 holes in the
Trocadero Museum, Paris. This museum is now incorporated in the Musée de l’Homme, Paris,
but K. C. Murray was unable to find any mancala III boards there, and on inquiry was told that
there were none. None of the books that I have consulted for Malagasy board-games mentions
any other than the mancala IV game, katra (8.3.1), and Mrs. Danielli had never seen a mancala III
game in the island.
8.1.5] MANCALA. PART II 207
half of the board.’ Mancala IV is probably of Arabic invention, since W. H.
Ingrams (Zanzibar, London 1921, p. 256) says that two of the three varieties
of bao which are played in Zanzibar are known as kiarabu, ‘Arabian’.
The mancala IV games fall into two distinct types. In one, all captures are
removed from the board and placed in the players’ stores. In the other type,
each player re-enters any beans that he has captured on his half of the board.
Presumably, therefore, all boards with stores belong to games of the first
type, as there is no need for stores in games of the second type.

(2) GAMES WITH CAPTURES REMOVED FROM THE BOARD

These games are played in Zanzibar, Mozambique, Nyasaland, North
and South. Rhodesia, the northern Transvaal, Damaraland, and by the
Hottentot tribes of South Africa. Broadly speaking, the peoples who play
this variety of mancala IV live farther south than those who play the varieties
in which captures are re-entered, but Dr. W. L. Abbott collected from the
Wa-Chaga tribe on Mount Kilimanjaro, Tanganyika, a board for Ochi (see
Fig. 85) which has 4x 6 holes and two stores (now in the U.S. Nat. Museum,
Washington). No description of how this game is played was obtained.

FIG. 85. OCHI BOARD FOR MANCALA Iv. Wa-Chaga tribe, Mount Kilimanjaro
(after Culin)

These games only differ in their initial arrays and in the rules regarding
captures. Singletons can only be moved when the player has no loaded hole
and then only to an empty hole. Moves may begin from any loaded hole
and a move may consist of several laps. The move ends when the last bean
in hand is sown in an empty hole. I classify the games by the method of
capture.
First form of capture. When a player’s last bean in hand is sown in an empty
hole on his front row, and his opponent’s opposite front-row hole in the
same file contains beans, or both of his opponent’s opposite holes, front and
back row, contain beans, these are taken.
8.2.1. Zanzibar: Bao kiarabu, ‘Arabian bao’, or kiarabu (W. H.
Ingrams, 256, where it is called ‘the true Arabic form’). 47 holes and two
stores. Two beans in each hole. Several laps to the move; one round; play
only anticlockwise.
A move may start from any loaded hole on the player’s rows. If the last
bean in hand is sown in a hole containing beans, these are lifted for a new lap;
if in an empty hole in his front row and the opponent’s opposite front-row
208 MANCALA. PART II [CH.8
hole contains beans, these are taken. If the opponent’s opposite back-row hole
also contains beans, these are taken as well.
8.2.2. Zanzibar: Bao kiarabu or kiarabu (W. H. Ingrams, 256). 4x8
holes and two stores. Three beans in each hole. Several laps to the move; one
round; play either clockwise or anticlockwise.
A move may begin from any loaded hole on the player’s rows. If the last
bean in hand is sown in a hole containing beans, these are lifted for a new
lap; if in an empty hole the move ends, and the player takes the contents of
either or both of the opponent’s opposite holes.
8.2.3. Mozambique and Nyasaland: Njombwa (Dr. M. G. Sanderson,
731, who says that the game was already obsolescent in Nyasaland in 1913
and in process of being replaced by bao, 8.3.11). 4x8, 4X9, or 4X 10 holes
and two stores, one at each end of the board. Three varieties are played, only
differing in the number of beans, their arrangement and opening play. The
descriptions are for the 4x 8 board.
(a) The players first check the number of beans by putting two in each
hole. The board is then cleared, and each player puts twenty-nine beans in
his left-hand hole on his back row (A, a), two in the next hole (B, b) and one
in the third hole (C, c). The opening play is
X’2B 'Xi2k Xoo Ko2K eb X GH OY 3b
Yoh ¥ 26. 6-20 ay 2k sob 3H 3e
A2C XX 2F eel ka 26An 3K 3h
Wioc =Ye2i a Yo2i 2N Y 29a 3k
X 2D X 2G X aj 2P 2n
Nad). Y 20" Ye) 3B 2p
giving the position
ag
CRE MN WT RG WME goa)
aye i OR ON eae re Le,
Sal RE cathe Spek lat ai ba
27 tore ae tgamow 37 ty ao
x

from which the play proper begins.

(b) Each player has sixteen beans, one in each hole. X then plays 1H, 2I,
2K, 2M, 20, 2A, 2C, 2E, transfers the two beans in G to H, and removes 2]
from the board. Y then plays the same succession of moves on his side of the
board, producing the position

ONOn’N
OHOOOoONON’N O’On1S)
toy
ten
fey bk
ONO

from which the play proper begins.

8.2.3] MANCALA. PART II 209
(c) Each player has twenty-nine beans which are arranged thus:

WN
y’oON
NN
be WN
NNN &boN
NN NAN
N ON
BN

All three games are then played in the same way as described on p. 207.
8.2.4. South-West Africa, Damara, Herero, and Hottentot tribes. Huts
or Hus, ‘holes’ (L. Schultze, Aus Namaland und Kalahari, 1907, p. 313).
4X16 to 4X22 holes according to Schultze, but a figure, which was pub-
lished in ILN. some years ago, shows 4X8 holes and no stores. Schultze’s
description relates to the 416 board. Two beans in each hole on the two
outer rows and on each player’s right-hand half of the two inner rows.
There are various forms of opening play; Schultze gives four forms: (1) X
transfers the two beans in M to U and V; (2) X transfers the two beans in I to
V and W;; (3) X transfers the two beans in X to M and N;; (4) X transfers the
two beans in K to W and X. In each case Y makes the same transfers on his
side of the board. The game proper begins from the rearranged position.
Play is only clockwise. The game is played in the same way as 8.2.1.
Second form of capture. This only differs from the first form in one particu-
lar. In addition to the capture of beans in the opponent’s opposite holes on
the same file, the player takes the contents of any other single hole on his
opponent’s rows, whether on the front or back row.
8.2.5. Nyasaland, Angoni tribe: Nchombwa or Nsoli (Sanderson, 733).
4X10 to 4X20 holes, each row containing an even number of holes. Two
beans in every hole except the player’s extreme right-hand hole in his front
row, which is left empty.
8.2.6. Nyasaland, Atonga: Nchuwa (Sanderson, 734). 4X6, 4X9, 4X I2
(the game described), or 4X15 holes (godi). Sanderson says that the 4x15
board gives the most interesting game. Two beans (machi) in each hole. For
the opening play, X lifts and sows the contents of his right-hand hole on his
front row (M), then the contents of the third hole in the anticlockwise
direction (P), and so for every third hole. Y then does the same on his half of
the board, producing the position
NG
Se a3 On BSi eS be Ole ese o3 Ol esi 3 tO
Ole Sa. Rk S Seek SR Ope eeSal? Soe OSES
3 eS et OS ie SO ar TS aM eS EO
OF Se Sa Oso nS eeeO re 3, Sa On MS) amrS
xX

X then transfers two beans from any of his loaded front-row holes to an
empty hole on his back row and takes the contents of the two opposite holes
5356 P
210 MANCALA. PART II [CH. 8
in the same file as the hole from which he transferred the two beans, and the
contents of any single hole, front or back row, on his opponent's side of the
board. Y then does the same, and the opening play ends.
8.2.7. Northern Rhodesia, Ba-Ila tribe: Chisolo or Natatu, ‘one of three’
(E. W. Smith and A. M. Dale, ii. 232-7). 4.x 7 (the game described) to 4 x 10
holes. Three beans (Iubwe, ‘pebble’) in each hole except the four front-row
holes on each player’s right hand which, reading anticlockwise, contain
0, 0, I, 2 beans. Each player begins by making one move (kubingula) clock-
wise. Each has the choice of making his second move either clockwise or
anticlockwise, but having chosen the direction, all subsequent moves must be
made in that direction. This may result (as in the following game) in the
players moving in opposite directions.
Smith and Dale record the following game on the 4x7 board from
native play:
X *3aL: 3m3b3a X *3F Xx e2C Nie
V. 31-3 M3 BgAl sYu4eeaR4Dil Vers tHib sro
X *3K : 3n3k X *3G DED) Y 2n
Y 2) : 3J3E2I Y -21 *2B:2h3e X*2M: 1k3e
X *3N : rk3dri 2k Seep ien Y 2a
Y3t:aM3C X78 X *2A X20 Silteab
and wins
This game was finished in twenty-one moves; with larger boards the game
may be much longer; one which Smith and Dale recorded was only con-
cluded in 117 moves. Occasionally the game may be lengthened by kuisha
Balumbo, ‘passing in foreigners’. When one player sees that he is losing, he
may add six or seven beans to his depleted holes. His opponent may, if he
likes, do the same, but unless he does so, the first player may not ‘pass in
foreigners’ in two games running.
8.2.8. Northern Rhodesia, Ba-Ila tribe: Chisolo (Smith and Dale, ii. 237).
4X7 to 4X10 holes. One bean in each hole. Since there are no empty holes
and all the holes are singletons, some opening play is necessary. X removes
the bean from one of his front-row holes and from both the opponent’s holes
on this file, and Y then does the same. The game proper then begins.
8.2.9. Northern Rhodesia, Ba-Ila tribe: Chisolo or Nambidi, ‘one of
two’ (Smith and Dale, ii. 237). 4x 7 to 4x 10 holes. Two beans in each hole.
X removes the beans from one of his front-row holes and the beans from
his opponent’s two holes in the same file and the contents of one other of
his opponent’s holes. Y then does the same, and the game proper begins.!
8.2.10. Nyasaland, Achikunda: Spreta (Sanderson, 736). 4X 10 to 4X20
holes, each row containing an even number of holes. One bean in every hole
except each player’s right-hand end-hole in his front row which is empty,
and the third hole from his left-hand end of his front row which contains
* A fourth variety of chisolo is 8.2.18. These four games are said to be the only board-games
played by the Ba-lIla tribe.
8.2.10] MANCALA. PART II 211
two beans. Play always begins by lifting and sowing the beans from the hole
with two beans.
8.2.11. Northern Transvaal, BaVenda tribe: Mefulhva (H. C. Stayt). 4x6,
4X8, 4X10...4%X28 holes, but usually from 416 to 4X28 holes, and
two stores, one at each end of the board. Two beans in each hole except

8A
SOE
SP,

\
SS
An a
\\,) x

FIG. 86. MEFUHVA. BaVenda tribe, N. Transvaal (cp.)

the end left-hand hole in each player’s front row which is empty, and the
next hole to it which contains one bean. Play anticlockwise. Only men play
and it is tabu to play after sunset for fear of cattle-raids, or to use fruit pips as
beans in the rainy season which would invite thunderstorms. Stayt records
the following game on the 4x6 board:
me 2ieee2ted (AF X 3A X 2H X al
Y 2j: 2G2F2E 3H 2D Y 2j 2K
X 2B Y2be -Y 2¢: 2013 al ViETE
BD riltaec, | 2G) Xe) aiciik” Ko: 2kibrawx 21
WE ANE API aes 1 Y id Wieii
> AC, 2K X 2G X 2A: Igic
and wins
8.2.12. Mozambique, BaThonga tribe: Tshuba or Shimunana (muna =
‘four’) (H. A. Junod, 345). 4X4, 4X8, 4X10, 4X16, 4X22 holes (but not
4X 8 because ‘successful combinations are impossible’, but cf. 8.2.11 which is
really the same game). Two beans in each hole. Played by men. Junod gives
the following game on the 4X 4 board:
X 2G Y 2e X 3B:3h3are = Y«re
3A 3g Y 1f:3F1C1G X 1F:1f
3D : 2bacef 3d:3H3E X 1E:1gid and wins
212 MANCALA. PART II [CH. 8
Third form ofcapture. This only differs from the first form in one particular.
In addition to the captures of beans in the opponent’s holes on the same file,
the player takes the contents of any two other holes on his opponent’s side of
the board, whether on the front or back row.
8.2.13. Nyasaland and Mozambique, Manyanja tribes: Msuwa (Sanderson,
735).4X 10, 4X 12...4X 20 holes. Two beans in each hole except the players’
two end right-hand front-row holes, of which the outer is empty and the
inner contains only one bean.!
8.2.14. Nyasaland and Mozambique, Manyanja tribes: Msuwa wa kunja
(Sanderson, 735). 4X 10, 4X12... 4X20 holes. Two beans in each back-row
hole, none in the front-row holes.
8.2.15. Mashonaland, Mashona tribe: Tsoro (C. Bullock, 203). The num-
ber of holes is not specified. Two beans in each hole.
8.2.16. U.S.A.: Chuba (Culin, c, 605). An adaptation of an East African
game, published by the Milton Bradley Company, Springfield, Mass.,
U.S.A., 4X11 holes. Thirty beans on each side arranged thus:
¥-
st
2
2
ONNH
“x HH mH
NA A
NN
H me
NA I NNA
He
& me
NNHW
NN
H es
OH
NHme
NY
O

xX

This differs from the other games with the third form of capture in the
following particulars: the extra captures must be taken from two holes in
the same file, and when a player is reduced to singletons, his opponent takes
them all.
Fourth form of capture. This only differs from the third form in that the cap-
tures are also made when the player’s last bean in hand is sown in an empty
hole on his back row.
8.2.17. Mozambique, BaThonga tribe: Tshuba (Junod, 349). 4x 8, 4x 10,
4X16, or 4X22 holes. Two beans in each hole. Junod’s account establishes
the method of capture, but does not go into detail. Schapura, Bantu-speaking
Tribes of South Africa, 1937, p. 127, names the game moruba.
Fifth form of capture. When in sowing the player’s last bean in hand is sown
in an empty hole on his front row and the opponent’s opposite front row
hole is empty but his back row hole on this file contains beans, these are
taken.
8.2.18. Northern Rhodesia, Ba-Ila tribe: Chisolo or Namudilakunze,
‘eating on the outside’ (Smith and Dale, ii. 237). 4x7 to 4X10 holes. One
* This game appears to be the same as the Mchombwa or Msua mentioned by O. Werner,
Natives of British Central Africa, 1906, p. 113, who cited Rev. D. C. Scott, Cyclopaedic Dict. of the
Mang’ anja, 1892. Scott says 4X 6 or 4x 9 holes: ‘ifAhas more beans than B in the opposite holes,
he takes them. A wins when B has no more left,’ but he obviously had no exact knowledge how
the game was played.
8.2.18] MANCALA. PART II 213
bean in each hole except the end left-hand hole in each back row which is
empty.
8.2.19. Nyasaland, Achikunda: Sute (Sanderson, 736). 4X8, 4XIO...
4X20 holes and two stores. One bean in every hole except the right-hand
end-hole on each player’s front row.
The following game differs in the arrangement of the beans, and the
method of capture.
8.2.20. Locality doubtful; possibly Tanganyika or Northern Rhodesia:
Chisolo (A. G. Shirreff, who obtained it from an English hunter whom he
met when travelling from the Cape to India). 4x7 holes and two stores, one
at each end of the board. Two players, each with twenty-one beans arranged
as follows:
Y;
OeSi) Ae 33 ae TO
On Om Okt OREO mmsO
Ol OM OM OR OMEEOMEEO
Ole Det2 5 3) eA SO
x

Several laps to the move. Play is clockwise. A move may start from any hole
on the player’s half of the board. If the last bean in hand is sown in a hole in
his front row and the opponent’s opposite front row hole contains beans,
these are taken, but if it is empty, the move ends. The player wins who first
takes all his opponent’s beans.

(b) GAMES WITH THE RE-ENTRY OF CAPTURES

These games, which include some of the most complicated forms of
mancala, are played in the southern Sudan by the Longarim, Topotha,
Lokathan, Zande, and Lotuko-speaking Lango tribes; throughout Uganda;
in Kenya, west and south of Lake Rudolf by the Dodoth, Didinga, and
Turkana tribes, and as far east as Mombasa; in Ussinja and Ukarere Island on
the Victoria Nyanza; in the Belgian Mandate in Ruanda and Urundi; in
Tanganyika by all the tribes of Bukoba Province; in Zanzibar; and within
the present century, by the Yao tribes in Nyasaland and Mozambique.
All the games are normally played on a board of 4x 8 holes without stores
by two players, each of whom has thirty-two beans, though occasionally
larger boards and more beans may be used. As a rule the board is marked
out on the ground, but cheap boards are common in Uganda (R. S. Shackell,
a, 14), and the Swahili and Baganda and other more advanced races often use
boards of beautiful make and finish (Man, 1931, xxxi. 132, 143).
The game is to capture or immobilize all the opposing beans. In some, but
not in all of the games, the game is also won if one player in one turn of play
takes all his opponent’s beans in the two end-holes of his front row.
The main difference between the various games are:
1. The discrimination of certain holes, usually either A, P, a, p, or A, B,
214 MANCALA. PART II [CH. 8
O, P, a, b, 0, p, which, following Sanderson, I call reverse holes.t Beans in a
reverse hole may be lifted and sown clockwise, if by so doing a capture can
be made.
2. The conditions governing captures. These involve the definition of
holes in opposition. Beans can only be taken when they occupy holes in
Opposition.
3. The rules for the re-entry of captures.
4. The opening play.
The following rules are common to all the games:
1. The game is played on one round, and a move may consist of several
laps.
2. No singleton can begin a move or lap or make a capture. It follows
that a player who is reduced to singletons is unable to move and loses the
game.
3. A move may begin from any loaded hole on the player’s side of the
board. If this hole is a reverse hole, the move may be made anticlockwise or
clockwise, but only clockwise to effect a capture. If it is not a reverse hole, it
can only be made anticlockwise. If the last bean in hand falls in an empty
hole, the lap and move end and the opponent plays; if in a hole which con-
tains one or more beans, and is not in opposition as defined for the game
being played, the lap ends and a new lap is begun by lifting the contents of
this hole and sowing them in the usual way. Further laps may follow until
the last bean in hand falls in an empty hole or a front-row hole from which
the rules permit a capture. Captures are then re-entered in accordance
with the special rules of each game, and the new lap so initiated is played
like other laps.
Notation. The re-entry of captures necessitates some modification of the
method described on p. 164. Moves in single brackets are replacements of
captures, thus (7k) means seven captured beans are re-entered in k and are
then sown in the usual way, i.e. in 1, m, n, 0, p, a, b. Moves in double
brackets are re-entries of captures which are not sown, thus ((4A)) means
that four captures are re-entered in A and left there. In bao (8.3.11), where
each player has twenty-one beans in hand which have to be entered, one in
each move, until the stock is exhausted, this entry is recorded by the notation
for the hole in which it is placed; thus, P means that a bean is entered in P.
When recording captures, I give the holes from which they are taken, and
the total number (if more than one) taken, thus 2k : EL(s5) means that two
beans are lifted from k and sown in 1, m, taking five beans in all from holes
E and L.
I classify these games by the number of reverse holes, beginning with
games with no reverse holes and continuing with those with two, four, and
more than four reverse holes.
8.3.1. Madagascar: Katra (R. Linton, 261, for the Tanala and Imerini
? Driberg called them turning bases.
8.3.1] MANCALA. PART II 215
tribes; Dubois, 844 and BAM., 1933, p. 261 for the Betsileo tribe; A. C.
Haddon in CP. for the Sakalava tribe; a board from the Mayungi tribe is in
the Musée de l’Homme, Paris, no. 41.17.29). Played by women and older
children only (cf. the Malay games). 46 to 4X 10 holes, but usually 4x 8.
The Tanala boards have one store ‘for spare pieces’; the Sakalava boards have
two stores, both at the same end of the board. There are no reverse holes.
Two beans in each hole. Several laps to the move; one round. The first move
determines the direction of play; if it is made clockwise, all moves are clock-
wise, if anticlockwise, anticlockwise (cf. again the Malay games).
A move may begin from any hole on the player’s side of the board, but
usually from a back-row hole. If the last bean in hand is sown in an empty
hole, the move ends; if in a back-row hole containing one or more beans,
these are lifted for a new lap; if in a front-row hole containing beans and the
opponent’s opposite front-row hole is empty or a singleton, the move ends,
but if this hole contains more than one bean, these are lifted and added to
the beans in the player’s hole in which he sowed his last bean, and all the
beans now in this hole are lifted for a new lap which is played in the usual
way. If the opponent has no beans in any of his front-row holes, captures are
made from his back-row in the same way.
A game may last for hours, but can be shortened by an agreement that
captures may be made simultaneously from both of the opponent’s rows.
This variety is called katra boaoaka by the Imerini, but the Tanala have no
special name for it.
Linton gives the following opening moves of a game on the 4x 8 board.
All moves are made anticlockwise.
x 2E Xs] oz) ok 3D agra P
3G: 2n 50 Y 2d 6i
The positions after X’s first move and Y’s first move are
2 QD 2» Da DB eB A aa Y FO YF © 2A za
Be 04. 2u 2) Be 216. 2 OT yA 3 os 35 aes
gh ah aye ae OP OS O73 tet oe toe somes
Saat os «Same A eS Sie sieoe Ome A wees

8.3.2. Kenya, Didinga and Dodoth tribes; Uganda, Karamojong tribes;

southern Sudan, Longarim, Topotha, Lokathan, and Lotuko-speaking Lango
tribes: Pereaiini (J. H. Driberg, “The Game of Choro or Pereaiini’, in Man,
XXVii, 1927, 114, 127). The description relates to the Didingas who learnt
the game from the Dodoths, a hill tribe living about two days’ journey to
their south, which had been driven by tribal wars to live in the plains and
remains a pastoral people despite the change in their environment (Driberg,
The Savage as he really is, 1929, p. 26).
4X8 holes (korok, ‘cattle kraal’), but the Longarim and Topotha tribes
sometimes enlarge the board to as much as 4X 16 holes. Two beans (tang,
‘cow’, pl. ten) in each hole. The back rows are mugur, ‘darkness’; the
216 MANCALA. PART II [CH. 8
front rows, kor, ‘sun’. There are no reverse holes and all moves are anti-
clockwise.
Four holes on the same file are said to be in opposition when one player
has his front-row hole and the other both his holes occupied, and the beans
in these two holes are said to be en prise. If in sowing, the player’s last bean
in hand falls in a front-row hole in opposition, he takes the contents of both
the opponent’s holes in this file which are en prise, and transfers his captures
to the empty hole from which he lifted and sowed his beans, and sows them
from this hole.t Any further captures in this turn of play are transferred and
sown in the same way. Thus, suppose X has six beans in H and one in N and
Y has two in e, three in f, two in k, and one in 1: X plays 6H : 3f2k, lifts the
five captures, sows them, starting from I (as though they were 5H), takes
three in e and one in | which are now en prise, and sows these three beans as
though they were 3H, &c.
Opening play. The two players play simultaneously, lifting and sowing the
beans from any of their back-row holes, usually from G or H, g or h. All
the rules of move and capture hold good with one exception: so long as no
bean has been sown in a front-row hole, the back-row holes are treated as
in opposition and captures can be made from them, but, once a bean has been
sown in a front-row hole, this privilege ceases and captures can only be made
from the front row. The opening ends when both players in sowing have
reached empty holes, and the first player to do this begins the game proper.
8.3.3. Uganda, Acholi tribe, Lango Province: Choro, coro (J. H. Dri-
berg, op. cit., 127; Capt. E. T. N. Groves, SNR., ii). 48 holes (kul). Two
beans (dyang) in each hole. The back row is dyewor, ‘darkness’, the front row,
cheng, ‘sun’. A, O, a, 0 are reverse holes (alok, ‘turnings’). The game only
differs from 8.3.2 in the rules relating to reverse holes. These are:
1. Whenever a player in his turn of play ends a lap in a reverse hole, he
may play the resulting lap (but only this one lap) clockwise, provided this
lap ends in a capture. Otherwise the lap must be played anticlockwise.
2. Ifa player begins a turn of play from a reverse hole, he may play it in
one of three ways: (a) anticlockwise in the usual way; (b) clockwise, pro-
vided thereby he makes a capture. The captured bears are transferred to the
reverse hole from which the move began and may be sown clockwise to
make a second capture, and so long as the resulting laps begin from the
reverse hole and make captures, they may be played clockwise; otherwise,
they must be played anticlockwise; or (c) when the player has exhausted all
the captures possible by clockwise moves from one reverse hole, instead of
playing the last capture anticlockwise, he may leave it unmoved in the
reverse hole and proceed to another reverse hole and exhaust all the possi-
bilities of clockwise moves from it, and, again leaving the last capture in this
reverse hole, he may start a lap from any other of his holes.
* In actual play, the captured beans are kept in hand and the player starts sowing them from
the next hole.
8.3.3] MANCALA. PART II 207,
Driberg illustrates these rules by the following composed position:
Omes LON? mee To tOMeT
i GO C23) G7 @
F@© 2 t BL © ©
So.0% OF 320" OU. One
X may begin his move from a reverse hole and play *5A : bm(6); *(6A) : °
cn(8); ((8A)); *2P : el(s); ((s0)), 4D; sH, &c., or from an ordinary hole and
play 4H: dm(6); (6H); 3N; *6A: ch(8); (8A) : ap(3), &c.
8.3.4. Tanganyika, Bugufi; Belgian Mandate, Urundi: Ikibuguzo;
Tanganyika, Kilaya tribe, Bukoba Province: Uruheso or Uruheisho (R. de Z.
Hall, CP.). 4 8 holes. Two beans in each hole. A, O, a, o are reverse holes.
Only differs from 8.3.3. in its opening play and in the addition of a second
method of winning by taking the beans in the opponent’s two front-row
end-holes (I and P, i and p) in a single turn of play (cf. 8.3.7).
Opening play. The players simultaneously rearrange their beans in one of
several conventional arrays, both adopting the same array. Only two are
given, as placed on the lower half of the board:
OL Og 3 VsOn Je O8' 30 1Omana AON ONG 20nd. Of0n 6
Opt OrsO250 G0; SOC O82, 444444 4 4
x x
The players then move alternately. Thus, after starting from the first of these
arrays, A lifts each group of three beans in turn, beginning from N and
placing two beans in O and one in P, then lifting 3L and placing two beans
in M and one in N, and so on. Or, starting from the second array, the
groups of four beans are lifted in turn and placed anticlockwise, two in the
next hole and one in each of the two holes beyond. If in these operations a
bean falls in a hole already occupied and that hole is in opposition, captures
are made as in 8.3.2 and 8.3.3.
8.3.5. Tanganyika, Bukoba Pr., Buganda tribe: Mweiso; Baziba, Bayoza,
Bahumba, Banyambo, Bakoba, Banyanangiro, Bazinza, Baha, Basumbwa,
and Banwani tribes: Ruweisho or Bahendagabo (Unsigned note, CP.).
4x8 holes. Two beans in each hole. A, O, a, 0 are reverse holes.
Opening play. As in 8.3.2, except that no captures may be made from a
back-row hole. The simultaneous opening moves usually begin from H, G,
or F (h, g, orf). It is forbidden to begin from D, d or E, e because ‘this would
result in the capture of more than half the opponent’s beans in one move’.
The game proper is played in the same way as 8.3.4.
8.3.6. Tanganyika, Basubi, Bahambazi, and Barundi tribes: name un-
recorded (unsigned note in CP.). 4x8 holes. Two beans in each hole.
A,B, O; P; a,b} 0; p are reverse holes,
Played in the same way as 8.3.5.
8.3.7. Uganda, Kampala (on the Nile, north of the Victoria Nyanza):
Mweso (R. S. Shackell, a, 14; b, 119). 4X8 holes. Four beans in each
of the two outer back-rows. A, B, O, P, a, b, 0, p are reverse holes. The
218 MANCALA. PART II [CH. 8
player wins who captures or reduces to immobility all his opponent’s beans,
or who takes in one turn of play all his opponent’s beans in I and P (i and p)
which necessarily means that he also takes all the beans in A and H (a and h).
These holes are called emitwe, ‘heads’, and this victory is known as nkutemye,
‘I cut your head off’.
Opposition. Four holes in the same file are said to be in opposition when
one player has his front-row hole and his opponent has both of his holes
occupied. The opponent’s beans are en prise.
Moves. Singletons cannot be moved. Otherwise moves are played in the
usual way and a move may include a number of laps. Clockwise moves
(okutebuka) may be made from a reverse hole, provided they end in a capture,
and following laps may also be played clockwise so long as each lap ends in
a capture.
Captures. Beans en prise are captured when the other player sows his last
bean in hand in his front-row hole in the same file as the beans en prise, which
must already contain one or more beans. The captured beans are then lifted
and sown, starting, not from the hole from which the capture was made, but
from the hole next to the one from which the lap or move began.
Opening play. The first player begins by rearranging his beans in his holes.
He may place any number of beans in a hole and leave some holes empty,
indeed, it is advisable to leave one hole in each file empty, so as to have no
beans en prise. Two favourite arrays are the following (shown on the lower
half of the board):
One On OM Om ONO MELON aT dar OmnOMm3 ns} mars EO
Td) (Gir 40 0-100 6 0 0 ‘0 fOr Old, 17
Xs x
Kyeso kyanyinye Kyeso kyansanya
(‘the 14 game’) (‘the 17 game’)

The ‘17 game’ is useful when the player aims at the nkutemye win. The
second player then rearranges his beans as he likes: it is not necessary to copy
the array adopted by the first player.
Mr. Shackell gives the following game which begins from the original
arrangement in which each player has four beans in each hole of his back row.
Opening play. X plays 4B; *4O; transfers two beans from A to O and two
from H to J; 2J; 2H.
Y plays 4h; *4a; 4g; 2k: EL(6). Since this is a capture by a legitimate
move, the opening play ends, but Y continues his move.
Y (7k) me GC) Y *30: DM(9) Masai)
sbt 8F (90) 6m
X 6C: ai(3) Y *2p7EL(3) 2h 7c
SPOsition?is 0 aplaans aSa Snes OME
D2 Oh YB 2 &
Ol 2} 80) “GY Gy ie Ge i
Ze 0 OG! 081 OO 0
8.3.7] MANCALA. PART II 219
Xe2E bo(3) = eX AgA SEk(s) ©X (2P) X 6L
(3H) (5A) 9B : cn(2) 2B
3K : £k(7) 3F : ap(z) (2B) 5D
(kK) (2F) 2D ‘a
Y 3g: BO(4) 4H OF 5M
(4g) 8L aL Y 2m: GJ(13)
X 3N 5D 3N (13m) : BO(3)
4A : el(12) sI 6A (3m)
(12A) IN 4G *p- FK(7)
4M 4E 9K (*7p) : AP(15)
Y 7d: CN(4) Y 2e 2D (Isp)
(4d) X 2H: bo(z) 2F 70
sh (2H) 5H 3f
X (61) :g}(7) 13J 12M 4i: EL(12)
5G tol (12i)
5P 3L ike, 4e
2E 50 4N X 3N
4G 3D 5B Y any
4K Y 81 4G X 3C and cannot play,
6O X 2N 4K and loses
2C AE Apa) 9O
vosn (*2P)2tk(2) 4H

8.3.8. Uganda, Entebbe. Mweso (H. J. Braunholtz, a, 131). This only

differs in two respects from 8.3.7. Moves from a reverse hole may be played
clockwise, even if the move does not result in a capture. Captures are not
placed in the hole from which the beans, making the capture, were lifted for
sowing, but in the hole in front of it, and the contents of this hole, both those
already in it and the captured beans, are lifted and sown.
8.3.9. Uganda, Mukonon: Mweso (W. G. Poole, manuscript description
in the Dept. of Ethnology, Brit. Mus.). Only differs from 8.3.7 in one
particular. Moves from a reverse hole are only allowed clockwise if they
end in a capture and the sowing does not extend to the back row.
Although mweso is now only played on the 4x8 board, it was formerly
played on boards of 4x6, 4X7, 4X9, 4X10, and 4x11 holes (see p. 160).
8.3.10. Uganda, Lango tribe: Choro (J. H. Driberg, 127). 4x8 holes.
Two beans in each hole. A, B, O, P, a, b, 0, p are reverse holes (aloka). Other
terms are dyank, pl. dok, ‘cow’; cheng, ‘sun’, front row; dyewor, ‘darkness’,
back row; awi, ‘hole’.
Opposition. The four holes on the same file are said to be in opposition
when one player has his front-row hole and his opponent has either his
front-row hole alone, or both his holes occupied, and the opponent’s holes
are then en prise. If the player, when sowing from a hole which is not a
reverse hole, sows the last bean in hand in a front-row hole in opposition,
he takes the opponent’s beans en prise, and sows them in his own holes,
starting from the front-row hole in which he sowed his last bean in hand.
220 MANCALA. PART II [CH. 8
If, however, he only took one bean, he adds it to the beans in the hole from
which he effected the capture, and sows them in the usual way. If the lap
which ended in the capture began from a reverse hole, the captured beans
are re-entered, not in the hole opposite, but in the reverse hole from which
the lap began.
Opening play. This is played as in 8.3.2, except that play must begin from
G or H (g or h) and beans are never en prise to beans in the back-row holes.
Moves. If the opponent has any beans en prise the player must begin his
move by lifting the beans from one of the opponent’s holes which is en prise.
If the opponent has no holes en prise the player may begin his move from
any of his own holes that are loaded.
Reverse holes. If a player begins a move from a reverse hole he may play it
clockwise, provided the move ends in a capture, and subsequent laps of
the move may also be played clockwise so long as each leads to a further
capture. Driberg composed the following position in illustration:
SOL OM ONO Lae OREO
Tecoma 7 ae OMnO
jie ete (6), Koy at chi Koy Xe
Buse ONO Om OREOREO
X to play

Y’s beans in c and n, h and i, j and m are en prise. X begins by taking 7m

and sowing them in his own holes starting from L and ending in B. He lifts
*4B and sows them clockwise in A, P, O, N, taking 3k. These he sows
clockwise in A, P, O and takes 2j which he sows in A, P, again clockwise,
and takes 4h and ri. As no more captures can be made by sowing clockwise,
he sows the beans anticlockwise in P, A, B, C, B, E.
8.3.11. Zanzibar and neighbouring coastal regions of Tanganyika in
which Swahili is spoken, Nyasaland, Yao tribe since 1900: Bao kiswahili,
bao, bau (Sanderson, 726 from whom the description is taken; W. H.
Ingrams, 257 from whom the native terms are taken). 4x8 holes (kishima,
pl. vishimo), and one store (kichwa, ‘head’) at one end of the board.! Two
holes, L and |, are square, and known as ‘villages’ (mji, musi) or kuu, ‘big’,
and special rules govern their use. The remaining holes are round. Two
players, each with thirty-two beans (komwe or soo, names of particular seeds
which are used as beans). One round; several laps to the move; play is some-
times anticlockwise and sometimes clockwise.
The players’ front rows are mbele, ‘in front’, and their back rows are nyuma,
‘behind’. The eight holes in each player’s front row consist of four reverse
holes, two at each end of the row (I, J, O, P, i,j, 0, p) and four optional holes
in the middle of the row (K, L, M, N, k, 1, m, n). Two front-row holes on
the same file are said to be in opposition when both are occupied.
Initial arrangement. Each player arranges ten beans on his front row,
? But a picture in ILN., 28 Sept. 1907, p. 460, showed a board of 4x9 holes.
8.3.11] MANCALA. PART II 221
putting six in L (1), two inJ (j) and two in K (k). Each player retains in hand
twenty-two beans which are known as namu.
The play may now be divided into three sections: the opening play, leading
up to the array which is favoured by the player, during which the rules of
move may be varied; the middle game (kunamua) during which the beans in
hand (namu) are entered; and the end game (mtaji) after all the beans in hand
have been entered on the board. If the players have no array that they favour
more than any other, the opening play may be omitted.
Opening play. The beans are arranged initially as given above, and each
player makes one move, during which the rules of the game may be dis-
regarded. X enters a namu in K, making its contents three beans which he
lifts and sows anticlockwise in L, M, and N, and his opening move ends.
N and k, M and | are now in opposition. Y enters a namu in k, takes the bean
in opposition in N and adds it to his three beans in k.! He then lifts 4k and
sows them anticlockwise in 1, m, n, 0, producing the position
Y
OORT Om On Om Om OMEO
OQ 2 O G 1 a i ©
Om OmOM TEN anOme LO
On1O OMOn Ono mOmO
x
in whichJ and o, 1 and m, N and | are in opposition.
Middle game. Each player has now twenty-one beans in hand, and each
turn of play is begun by entering a namu in a hole on the player’s front row
which already contains at least one bean, and if the player has any holes in
opposition, one of these must be chosen, and the contents of the opposite hole
are captured. This opposite hole is necessarily either a reverse or an optional
hole. If it is a reverse hole, the captured beans are sown from the same end of
the front row, e.g. if the captured beans were taken in 0 or p, they are re-
entered in I and sown towards the middle of the board, and if they were
taken from i or j, the re-entry is in P and they also are sown towards the
middle of the board. If the capture was made from an optional hole, the
re-entry may be made from either end-hole of the front row, and if more
than one bean was taken, they are sown from either end of the row towards
the middle of the board. If there are no holes in opposition in the front row,
the namu is entered in any front-row hole containing one or more beans,
and the beans in this hole are then lifted and sown in either direction. If all
the player’s front-row holes except the village (L, 1) are empty and the village
is not in opposition, the namu may be entered in it, but only two beans may
be lifted from it for sowing.
If the last bean of a sowing, whether of captures or of a lap, ends in a hole
in opposition, a fresh capture is made and the captured beans are treated as
1 Here the rules are disregarded. N is an optional hole, and beans taken from an optional hole
should be entered in a front-row end-hole, i or p.
222 MANCALA. PART II [CH. 8
above; if it is sown in an occupied hole, other than the village, the contents
of that hole are lifted for a new lap; if it ends in the village and the village
is not in opposition, or if no beans have previously been removed from the
village, the player has the option of ending his move without lifting the con-
tents of the village for a new lap; if the last bean in hand is sown in an empty
hole, the move ends.
End game. Moves may begin from any loaded hole in either the front or
back row. The contents of the hole may be sown in either direction. If the
last bean in hand when sowing a lap does not fall in a hole in opposition, no
captures can be made in that turn of play although further laps may be
played.
A player loses the game if all his front-row holes are empty or if his front-
row only contains singletons.
Dr. Sanderson gives the score of a game which was played by two Yaos.
It begins from the position diagrammed above, and I diagram the position
every few moves.
X L:m 2p Ap X 2K
P 3m I Y iM
Y¥ 022] XxX. V?2m 31 1
p, 0 |Fl| 56 | DN |
730:M. Y.o2J Pp P
p p ene *3P
O*0 6) (0.010. O40
Or .2) Tao or Gr 3) ot
OF foe IT, 07, 076
615670 otto oO “Oo fo
Yi:M XRaP eri). oy. pk Xo
1 PLOmer pP, 0:2) Yours
w O33} *5F p, 0 p, 0,n:2K
O, P *2D 50 p,0
3P 225 XO:j ap2L
*2V4 *P:k P Pp
P Pp *oP akc 7p
01070 6 oO .6 Ot;
‘ee Ce See leermae dt bye
TS ONO On Onno ero
200 ate Oe) Ome
X N:k XxX Nek Y 2b x *3P
Pea P xX Nik Ym
P era P *6m
#30 iZP 4 *2¢
Y j:60 i *5j *2¢
ii 31 X Pri "AE
sn él P *3p
8.3.11] MANCALA. PART II 223
2h Os 2020 12200 Teas
Ti ota TU 1A ka 0
O211 (iia t2O P05 Onto
Bai Oul2 a)Op 200562 32
X N:k X K:2n Vop;io Y *3p
I 1 *8p *on
Yr M YO, Xx On3p X N:k
i pil P,O,N:k P
21 p Ee er
& Nek 3p P al
I *3m vase} X N:k
2i XN? YaromM Pit
Yoz] I pes P
p Yn:2K p BE
7 ee Amer Wale Sia Seats Ras |
Ont JO0)0 Sas uo
025519 st) 07 oF 0°60
2. Ona AOU SG) eee
Yijs5O xX N:2k X 4H Vushvet
1, jas an P,O 2L Lado
X N:k Yi:P 2N Lik
P i 4A 2l
Y1I:M * 34 Yo *ad < 2f 4B 3k
i:P *] p, 0 P, O, N
i *on 40 *3N : 8n
*3i XN:k X 4Gin P:4q
bab Pp I: sp P,O,N,M
P20 Yj:0 LJ...M:1 %*2M
*4a i I eK
*2e X 22N 4l *3D
*39 2D 4M *2A 25p
54 2F 6A P, O70 7m
and wins
OS O84) Din OlrS Ace
0.26) 0 Oi, 0420 10m 0
TS 2 Ta On eed
Ql i Of @) WH A Bp

(c) CHILDREN’S GAMES

Dr. Sanderson adds two games played by the children of the Manyanja
tribe, Nyasaland, which he suggests are the probable ancestors of the adult
games of mancala IV. This I doubt, in view of the evidence for an Arabian
parentage. The children’s games have nothing, except the implements of
play and the operation of sowing, in common with the adult games, and
look to me as being the kind of imitation of the adult games that children
would make. The adult Manyanja dismisses them as wa chabe, ‘useless’ or ‘of
no interest’.
224 MANCALA. PART II (CH. 8
8.4.1. Nyasaland, Manyanja tribe, children: Chiana wa kunja (Sander-
son, 734). The board has four rows of holes, the two outer rows alone being
used. Four beans in each hole on these rows. The players own no holes by
right of position, but own beans and any hole may be regarded as belonging
to a player so long as it is occupied by his beans. Every move is begun by
lifting the beans of the player’s rearmost hole and sowing them anticlockwise
in the usual way, and consists of one lap unless his two front holes contain
one and two beans respectively, when he is allowed to play another lap. If
in sowing a player reaches holes occupied by his opponent, these holes and
the beans in them pass into his possession. If his rearmost hole is a singleton,
he first transfers this bean to the next hole and then lifts its contents and sows
them. If a player is reduced to a single bean, he moves this bean two holes
each move, until he has overtaken the opponent’s beans and made a capture.
The game ends when one player has lost all his beans.
8.4.2. Nyasaland, Manyanja tribe, children: Chiana wa bwalo (Sanderson,
735). The board has four rows and each player owns the two rows nearer
him. Four beans in each hole of the outer rows. Moves are made as in 8.4.1,
but beans are taken differently. Whenever a player’s beans enter his front
row and occupy a hole or holes directly opposite to occupied holes in his
opponent’s front row, he takes the contents of these holes and places them
in his store. The game ends when one player has lost all his beans.

(d) GAMES OF WHICH WE HAVE NO DESCRIPTION

8.5.1. Hanzoan: Morahha (Hyde, 232). 4x8 holes.
8.5.2. Sudan, Zande tribe: Banya (Major P. M. Larkin, 27). 4x 16 holes.
‘A kind of draughts, played by young and old, rules somewhat complicated.’
8.5.3. East Africa, A-Kamba tribe: Mutingwano (C. W. Hobley, a, 55).
‘The A-Kamba play the African game of bau or mutingwano as they call it;
instead of three rows of holes they play it with one row of ten holes. The
pebbles are thrown up and caught, if any are dropped in the process the
player loses points.’ Played in the month thandahi when crops are ripe.
8.5.4. Uganda, Turkana tribe: Ingilith (E. B. Embley, 195).
8.5.5. Uganda, Labwor tribe, Karamoja: name unrecorded (E. J. Way-
land, a, 196). 4x8 holes. ‘The game known as bau in Kenya, mweso in
Uganda . . .is also played.’
8.5.6. Tanganyika, Wa-Chaga tribe, Mount Kilimanjaro. Ochi (Culin,
c, 602 and pl. IV). 4x6 holes and two stores, both at one end of the board.
See Fig. 85.
8.5.7. Tanganyika, Barundi tribe: Ikisoro (H. Meyer, 66).
8.5.8. Congo Basin, Bushongo tribe, name unrecorded (Torday and
Joyce, 97). 4x 7 holes. A board from the Isambo clan is in the British Museum.
8.5.9. Comoro Islands: M’raantsou (Avelot, b, citing J. Repiquet, ‘Le
Bas Cavally frangais’ in Bull. Com. Afr. frang., 1903, Suppl. no. 11, 283).
8.5.10. Madagascar: Fifanga (Avelot, a, citing H. d’Escamps, Hist. et
8.5.10] MANCALA. PART II 225
Géog. de Madagascar, 1884, p. 433 and Corré, in Bérenger-Féraud, Les Peu-
plades de la Senegambie, 1879, p. 328). According to d’Escamps, the only
Malagasy game and played by both men and women. Nuts or seeds are used
as beans. Mrs. Danielli told me that she had never seen or heard of this game.
8.5.11. Mashonaland: Isafuba (Culin, c, 603, citing Bent, b). “The
mysterious game of the Makalangas, with sixty holes in rows in the ground.
Ten men can play at this game, and it consists in removing bits of pottery or
stone from one hole to another in an unaccountable manner.’

5856 Q
 7
THE, DISTRIBUTION AND ORIGINGO?
BOARD-GAMES
Tue catalogue of the board-games which I have collected is now completed,
and it shows that all that we really know about most board-games is where
and how they are played at the present time. In the case of the peoples who
have no history this is inevitable, but we are not much better off with the
more civilized peoples who have learnt to record their achievements, their
struggles with nature and other peoples, and their advances in standard of
life. In comparison with these, the adoption or invention of a board-game
plays a very small part, and would naturally be one of the last of the activities
to be recorded. No contemporary annalist recorded the arrival of chess in
western Christendom, or the introduction of the changes which resulted in
the modern games of chess and draughts. Indeed, as I have shown elsewhere,
the medieval writer thought it beneath his dignity to mention such trivialities
as the games played by his fellows, so—apart from merels, tables, and tafl of
the older games, or draughts and rithmomachy of the games that were
invented in the Middle Ages—we know very little of the histories of games
in the usually accepted meaning of the word, that is at all comparable with
what we know of the history of chess. For most board-games we have to
interpret their history from casual references and allusions in the light of that
of chess and tables, and on the assumption that all board-games have under-
gone a similar process of development.
That this method of procedure is not unreasonable is supported by the
vitality of board-games. As Sir John Myres (The Dawn of History, 9) has
pointed out, the greater part of history is not the record of things immutable,
but of change, and we know also that among savages and barbarous peoples
there is the least room for change in their way of life. In their case, at least,
we may assume that their games are unlikely to have been greatly changed
and that they will generally retain their primitive forms. Among more
civilized peoples, however, we must expect that their games have undergone
an evolution which may result in great changes of rule. The history of chess
in different regions shows this: the modern chess of Japan is very different
from its parent in China, and both differ greatly from our chess, though the
common ancestry of these games is certain. So long as the board-games of a
people are handed on from one generation to the next by word of mouth or
mere imitation, changes may easily occur: it is only when the rules of a
game are reduced to writing that they tend to be fixed. Changes of rule do
not destroy a game; they act as a tonic to a sick game and restore it to health.
 THE DISTRIBUTION OF BOARD-GAMES 227
Freiherr v. Holzhausen (Handbuch des Schachspiels, 8° Aufl., Leipzig 1912,
p. 26) argued that ‘a board-game which depends on the skill with which it is
played can only live if a small error leads to the loss of the game. On the
other hand it is irretrievably destined to death if a great majority of games
end without result, ic. in a draw’, but this is not true, or our draughts
would have died long ago. In six international matches played between
1884 and 1905 in which 1,181 games were played, 831 (71 per cent.) were
drawn. This did not kill draughts; it only stimulated players to look for a
remedy, which was found in the introduction of a new rule for match games
—‘the two move restriction’ by which the free choice of openings was
replaced by drawing lots for the opening moves—and draughts is played
more widely than ever. The fact is that the enjoyment of a board-game does
not depend only on its result. The smaller merels (3.3.1) has been played for
four or five thousand years although it is easy! to discover that the first
player has a forced win. All that happens is that the player who discovers this
ceases to play that game and turns to another, but children and primitive
peoples like the Ainus of Japan still play it, and the game continues to live.
The only game of the Middle Ages that has perished entirely is the highly
artificial arithmetical game of rithmomachy which was played in Europe
for more than four hundred years. It was the dullness and difficulty of this
game that killed it. The chances are that somewhere or other any board-
game that ever existed is still played, though possibly in a more developed
form. The only thing that may lead to the disuse of a board-game is the
competition of a better game, or a game of another kind which gives greater
enjoyment or satisfies other demands, such as whist in the eighteenth century
in France and England or bridge in our own day.
The catalogue of board-games in the preceding chapters contains some
270 games which are itemized by the countries for which they are recorded
or the peoples by whom they are played.? That many of the items relate to
the same game is obvious, and these are listed seriatim after the leading game
of the type. Other games are so closely akin that they can be regarded as
descendants of a common original, and these follow immediately after the
entries relating to this original. When discussing the history of board-games
and their development, I treat all these games as a single family and this
reduces the number of families materially. Thus, all the alquerque games of
Europe and Asia are treated as modifications of a single original, alquerque
(4.2.1), although the shape of the board and the number of men may differ
considerably. The same procedure is adopted for the existing games of
draughts (4.3.1), for the descendants of the Roman game, alea (2.8.3 and
T Oris it so easy ?It was known to the author of Alf. in 1283, yet Kraitchik (Mathematical Recrea-
tions, 1948, p. 290) still asserts that when properly played the game must be drawn, although he
actually gives the forced win.
2 Asarule in the catalogue I treat Europe as an entity without specifying the different European
countries, and include in it all those parts of the world which have been colonized by Europeans
who retain their European language.
228 THE DISTRIBUTION AND [CH.9
6.1.1), and for other groups of games. In this way I reduce the number of
separate board-games to some fifty or sixty, of which about half are played
by a single people. These are generally simple or trivial games, though they
include one outstanding game, the fanorona (4.6.1) of Madagascar. The other
half are played with various modifications by a number of different peoples.
Their distribution is shown in the table at the end of this chapter.
How are we to account for the fact that two peoples inhabiting widely
separated countries play the same game? There are two possible answers: one
is that the game was independently invented by the two peoples; the other
is that a game, like other inventions and ideas, tends to spread from the
people who invented it to their neighbours, and from them to other peoples
in the same way until it reaches the second people. Clearly a game must have
been invented somewhere, and presumably to meet some felt need, but it is
unlikely that different peoples would meet that need in precisely the same
way, and we can only accept independent invention in the last resort when
other explanations are impossible. The history of the games treated in
Chapter 2, and the history of chess, show that games have in fact spread
from one country to another, and that this is in general the right answer to
our question.!
The actual course of diffusion of a game may be complicated. The original
stage may be to any neighbouring people with whom there is contact for
exchange of goods or ideas, and these peoples may modify and improve
the game and become a new centre from which the reformed game is
diffused, and this diffusion may carry the reformed game back to its original
home. Thus the Parsi and Muslim games of chess have returned to India to
compete with the parent Indian chess. Or a game may reach a country from
two directions as in Siam, where chess arrived both from India by way of
Burma and from China.
The actual course of diffusion may also be blurred by gaps in our informa-
tion. The three-in-a-row games of nine holes (3.1.1) and the smaller merels
(3.3.1) are both played in precisely the same way in many parts of the Old
World which are separated by intervening areas in which these games have
not been noted. Both are very old games and it is possible that in the inter-
vening countries they may have escaped the attention of travellers who were
more interested in other phenomena. It is only recently that they have been
noted in Madagascar and in the Gold Coast and Nigeria. Or we may know
nothing about the ethnology of the intervening countries.
The immediate source from which a people has obtained a game is most
satisfactorily determined when the name of the game is adopted with the
game itself. Thus the route by which chess reached Europe from India is
established by the catena, Sanskrit chaturanga, Middle Persian chatrang, Arabic
shatranj or shitranj, and the Spanish ajedrez and Portuguese xadrez. This,
* I am dealing here with the history of an existing game, not with the invention of games in
general with which I shall deal later in this chapter.
 ORIGIN OF BOARD-GAMES 229
however, seldom happens. Most peoples give a game a new name from their
own language, often using the name of the seeds or beans which they use as
pieces, but occasionally attaching to the new game the name of another of their
games, as when Abyssinians call the larger merels shah, or the peasants in the
country surrounding Nuremberg call merels schafzagel, both meaning chess.
The existence of the following centres of diffusion in the Old World,
primary or secondary, may be established:
1. The lower river valleys of the Nile (Egypt) and the Euphrates-Tigris
(Mesopotamia, the ancient Sumer and modern Iraq) which were both homes
of ancient and early civilizations. In both regions what are essentially the
same games were being played in the third millennium B.c. These games
probably included games of alinement, war-games, and race-games, the
evidence being most complete for race-games since the permanent boards
used for these were treasured and buried with their owners and so have
survived to the present time. I know of no surviving boards for games of
alinement or war-games from Sumer, but boards have been found in Egypt
and Palestine which are incised on slabs of stone, and this may account for
their non-survival in Mesopotamia where stone was hard to come by, and
the boards for war-games may have been usually traced in sand or soil as
required. All these games spread to Palestine in the second millennium B.c.
and to the Mediterranean islands and Greece in the first millennium B.c. We
do not know enough of the relationships between Egypt and Sumer in these
early times to say in which country the games originated, but the course of
diffusion westwards is clear. One game at least also spread to Assyria and
possibly to the Hittites.
2. Northern India, where the first evidence for the existence of board-
games dates from about 500 B.c. The earliest diffusion from northern India
was to China from Kashmir, and possibly occurred in connexion with the
spread of Buddhism. It was only after the invention of chess that we have
any evidence for the interchange of board-games with Persia. Board-games
seem to have reached southern India by sea rather later, but we know too
little about the early relations between southern India and the west to be sure
of this, though the alquerque games seem to have been introduced from the
west and passed on to Ceylon and Indonesia in the early centuries A.D.
3. China had already obtained games of alinement by the time of Confu-
cius, but the creative period for games, which culminated in the invention of
wei-k’i about A.D. 1000, began much later, and the first diffusion of games
from China was northwards to Korea and Japan. Since the rise of the
Manchu Dynasty about 1600, emigrants from China have introduced Chinese
games into Siam and Malaya.
4. Greece acquired board-games from Egypt in the course of the first
millennium 8.c. and, probably through its colonies in Sicily and Italy,
passed these games on to Rome, and after the arrival of Christianity, some of
these games were carried to southern Russia.
230 THE DISTRIBUTION AND [CH.9
5. Rome. The games of latrunculi, duodecim scripta, and merels were carried
by Roman legionaries to all the frontiers of the Empire and adopted by the
local populations of the frontiers and so reached the neighbouring peoples.
In this way the Germanic tribes learnt of the existence of board-games.
Since they called them by the Roman name, tabula, it is probable that they
had not previously played any board-game.
6. Scandinavia, where, as the result of the Germanic wars and contact
with Roman legionaries, the Norsemen had obtained the idea of playing
board-games, the game of tafl was invented in the Roman Iron Age. The
Norsemen carried this game to Britain, where it was learnt by the Anglo-
Saxons from Danish settlers and the game was adopted by the English,
Welsh, and Irish.
7. The Arabs, who had already in Muhammad’s day acquired a number of
board-games, became active disseminators of these games to the rest of
Asia, Europe, and Africa. Despite the opposition of all the schools of juris-
prudence to the playing of games as prohibited in the Qu’ran, they were
played in the courts of the caliphs and by all classes, and were carried to all
the lands to which Islam extended. The adoption of chess, after the conquest
of Persia, led to the invention of many new games, for the most part deriva-
tives from chess and played on larger boards, but including some original
games like al-falakia (6.8.1) which had a temporary vogue in other Muslim
countries and in Christian Spain. The Arabs gave to Europe chess and the
use of two dice instead of three in tables, and helped to preserve some
Roman games through the Dark Ages, giving a new name to merels in
Castilian Spanish. They gave to Africa mancala and to southern Asia mancala
and the alquerque games.
8. After the discovery of the New World, European colonists and settlers
took their simpler board-games with them and passed them on to the native
peoples with whom they came in contact. The Spaniards took these games
to Mexico and taught them to the native American peoples, as is evidenced
by the occasional survival of their Spanish names. Other European settlers
have introduced draughts and fox and geese into Canada and some of the
Pacific islands, missionaries being often the intermediaries.
All these centres lie in the Old World, and the oldest of them all lie to the
north of the Equator. So far as the evidence goes, there is nothing that is
inconsistent with the view that board-games were invented first by the
Egyptian and Sumerian peoples, and that the idea of playing board
games with some of their games was passed on to other Asiatic peoples in
the course of the barter of goods, and finally reached China before 500 B.c.
when the Chinese were playing the same games of alinement that were
played by the early Egyptians. If this is true for China, there is no reason
why it should not be true for India, although the actual Egyptian games are
not known to have been played there, native games of the same class replacing
them. It has been suggested by Sir H. Johnston (see p. 158) that a similar
 ORIGIN OF BOARD-GAMES 231
diffusion of games may have taken place southwards from Egypt, but there
is little evidence in support of this view; the absence of any race-games or
games played with dice in Africa is significant, and most of the existing
African board-games were introduced with Muhammadanism.
So far I have been dealing with the evidence for the diffusion of board-
games in the Old World, and after the discovery of the New World to
America, and have shown that it provides a reasonable explanation for the
wide distribution of these games. But board-games were already played in
America before the time of Columbus, and since these have been used as
evidence for an Asiatic origin of American culture, it is necessary to examine
the validity of this claim which was first made by E. B. Tylor in 1878.!
Tylor, who wrote before much work had been published about the exist-
ing games of the North American Indians, based his argument on the
accounts of the game patolli (6.7.1, Fig. 68) in early Spanish histories of the
conquest of Mexico, and its similarity to the Indian game of pachisi (6.4.6
and Fig. $9), affiliated to chaupur (Skr. chatush-pada, 6.4.1), which was played
in India in the early centuries of the Christian Era, and is said to have been
introduced into China, A.D. 230-68, under the name ¢’shu-p’u (see p. 36).
Both games were played on cruciform boards with primitive forms of lots,
in India cowries or long dice, in Mexico staves. In chaupur and pachisi each
arm has three rows of cells on each of the four arms, and in patolli two rows
of cells on each arm.? In pachisi certain cells on the three rows of each arm are
crosscut, and the crosscut cells are cells of safety, men occupying a crosscut
cell being immune from capture; in patolli no cells are crosscut but the
boundaries of some cells on the arms are thickened, apparently to distinguish
the outer cells from the inner ones. In pachisi the men are entered from the
central square on cells on the middle rows and circulate round the outer
rows of the four arms and back again to the middle row on which they were
entered and are borne from this row; in patolli we do not know how they
were entered, but the analogy of other American Indian race-games suggests
that men were entered on cells at the outer ends of the arms, circulated round
the arms of the cross and were borne from the same cells in which they are
entered. It would seem accordingly that the resemblance between pachisi
and patolli is reduced to the fact that both games are race-games played with
lots on a cruciform board and nothing more, and that this is insufficient to
support any relationship between the games.
Moreover, there are other difficulties. The Chinese references to t’shu-p’u
do not suggest that its life in China was more than an interlude in a series of
changes which ended in the adoption of shwan-liu, and this alone was intro-
duced into Japan. The main course of diffusion of pachisi and other Indian
1 E. B. Tylor, ‘On the game of Patolli in ancient times and its probable Asiatic origin’ in JAI.,
viii (1878), 116 and ix (1879-80), 23; and also ‘On American lot-games as evidence of Asiatic
intercourse before the time of Columbus’ in IAFE., ix, Suppl. (1896), 3.
2 In totolospi (6.7.9) the Hopi Indians occasionally use a cruciform board in which each arm
has only one row of cells.
232 THE DISTRIBUTION AND [CH.9
games on the cruciform board was not northwards but west to Persia and
east to Burma, Malaya, and Indonesia. If diffusion of games extended to
America, it must have been by way of the Behring Strait, and there seems
no parallel evidence of an early diffusion of culture from India to the Asiatic
shores of this strait that would support the diffusion of pachisi thither. The
latest view of transfer of culture by the Behring Strait to the Eskimos of
Alaska is that it is not primeval, and its most prominent feature is in relation
to the weapons and methods used in whale fishery.*
The comparative study of the existing board-games of the North Ameri-
can Indians also suggests that they are a native invention, and that it is only
since the discovery of the New World that any diffusion from the Old
World has taken place. There is a good deal of evidence to suggest that the
native American games were originally divinatory, and some are still used
for divination. But there is none that the Asiatic games were divinatory in
origin.
There are two American war-games, awithlaknakwe (4.1.14) and totolospi
(4.1.15) which employ the interception capture that is used in the oldest war-
games of the Old World, but instead of the orthogonal move which is
characteristic of these Old World games, both use the diagonal move of a
single step that only appears in those medieval European games which were
transferred from a lined board (on which moves could only be made along
a marked line) to a latticed board. Culin (d, 877 and 879) first treated these
as native American games, but later (g, 785, 799) included them among the
games which were learnt from Spanish settlers after the conquest of Mexico
without any explanation why he did so, disregarding the evidence which he
gave in his first work for the antiquity of their game-materials, and the
absence of any similar game in the Old World. In this, I think he is entirely
wrong.
Ihave already (p. 150) described the general characteristics of the Ameri-
can race-games of the present day. They are usually played on hollow circular
or square boards, the empty middle containing a piece of rock or stone on
which the staves which are used to give the moves have to bounce. There is
no parallel to this in any games of the Old World, and none for the differen-
tiation of one stave by notching it, which affects the interpretation of the
throw. The track is usually divided into four quadrants or sides which are
separated by wider intervals known as ‘gates’ or ‘creeks’. These serve to
mark the terminals of the tracks for the different players or sides. Occa-
sionally they are points of danger in that a man which occupies the creek has
* See E. Birket-Smith, ‘Recent Achievements in Eskimo Research’ in JRAI., lxxvii, part ii
(1947, but only printed in 1950), 145-54, where he sums up the position thus: ‘It is clear that
Eskimo culture is deeply rooted in the old World. That does not necessarily mean that the Eskimo
as a people migrated from Asia to America. . . Ifwe may assume that the proto-Eskimo lived as an
inland people’ (which he believes) ‘near the timber line from Alaska to Hudson Bay, a continuous
contact across the Behring Strait is all that is needed to explain the subsequent development of
what we now call Eskimo culture.’
 ORIGIN OF BOARD-GAMES 233
to return to its starting-point and begin the track again. In both the form of
the board and this treatment of the gates or creeks, the games have no
parallels in the Old World. I think, therefore, that we are compelled to
regard these American games as purely of American invention and to accept
in their case independent invention since any other explanation seems
impossible.
This brings me to my last question. How did board-games come to exist,
and what were the conditions that led to their invention? As Karl Groos (The
Play of Man, tr. E. L. Baldwin, 1901, p. 196) has said, ‘it is difficult to find a
perfectly satisfactory answer to the question of their origin’. Perhaps one
difficulty is that scholars have looked for a single source whereas board-games
may have originated in a number of ways. Since their invention must have
been one of the earliest stages in the development of culture, rooted in the
conditions of life in their oldest and simplest phases, long before the begin-
nings of any recorded history, we can only conjecture how they came into
being, and the guesses which have been made do not commend themselves.
Most previous writers have assumed that the oldest games were athletic
games, and have in the main concentrated their attention on these. When
athletic games are mentioned in ancient history, they appear as an essential
feature of sacred or magical rites for the worship of the gods, for the fertility
of the race and the soil, and in connexion with funeral ceremonies. In all of
these the adult males of the tribe or people took an active part, and it seems
a natural and easy step in the course of time for adults to come to repeat these
games for pure recreation and enjoyment. So it is generally accepted that all
the ancient athletic games are secularized and degenerate survivals of magical
or religious practices, although there are differences of opinion as to the exact
nature of these practices, Haddon specifically excluding practices related to
divination, while Culin confines them to those that were divinatory. It is
often assumed that this must also be true of board-games. There is, however,
a gap between the athletic games that were played by large groups of men
and the sedentary games that are confined to a few players, which I find it
difficult to bridge. It is difficult to see how the private operations of the
magician could be adopted by the secular members of a tribe. I think that
we must look elsewhere for the origin of most board-games.
Most writers on board-games draw a distinction between the sources of
games of chance and games of skill and assume that the former are necessarily
the older. We have seen that the oldest surviving game-boards from Egypt
and Ur (though not from Palestine) were used for games of chance, but I
have already shown that this evidence is indecisive because games of skill
were more often played on boards marked out for the occasion on the
ground, and games of alinement must from their spread have been very old.
It was Culin (d, 679) who was the first to claim that board-games and card-
games are derived from games of chance; he says:
Games and playing-cards are both regarded as derived from the divinatory use of
234 THE DISTRIBUTION AND [CH. 9
the arrow, and as representing the two principal methods of arrow-divination. . . .The
basis of the divinatory system from which games have arisen is assumed to be the
classification of all things according to the Four Directions. This method of classifica-
tion is practically universal among primitive peoples in Asia and America. In order to
classify objects and events which did not in themselves reveal their proper assignment,
recourse was had to magic. Survivals of these magical processes constitute our present
games.
Later in the same work (d, 853, 858) Culin sketches the evolution of all
board-games from the Korean nyout (6.5.4) to chess and draughts, although
his view was admittedly based on the Cox-Forbes pedigree of chess (HC. 46)
which has been abandoned by all scholars since A. Weber exposed its falsity
in 1872-4.
Culin (g, 45) is also responsible for the view that race-games were invented
to keep a record of the successive throws of lots by means of counters shifted
along a track, the length of which corresponds to the winning score.! This
view is repeated by G. Marin (‘An ancestor of the game of ludo’, in Man,
Ixii (1940), 64). But in none of the Indian race-games which he describes, nor
in the American Indian games given by Culin, is the purpose of the game
merely the summation of a number of throws of the lots. At any moment of
these games a player’s total may be cancelled if he reaches a point already
occupied by his opponent, or otherwise complicated by the existence of
crosscut cells. It is precisely in these additional hazards that the charm of these
race-games consists.
Nor can I accept the theory of the origin of race-games which K. Groos
makes when he says:
There is but one opinion as to the origin of games of pure chance—namely, that
they grew out of the serious questionings of Fate in the form of oracles, and colour is
given to the theory by the custom of jesting with the oracle.”
The appeal to the oracle seems to me altogether different.
The only suggestion that I have found in previous writers which seems to
me to be helpful, is made by Groos and relates to war-games. He says:
The primitive races, who find it difficult to convey their thoughts in speech,
naturally take to marking on the sand, and hence the figures (i.e. game-boards) might
arise. If the leader of one of the more intelligent peoples wished to instruct them con-
cerning some part or future combat, it would be a simple method of illustrating his
meaning to draw an outline on the ground and represent the position of the hostile
forces by small stones or similar objects, whose movements would symbolize the
manceuvres of the forces or the advance of knights for single combat. This would,
no doubt, be exceedingly interesting to those conducting it, and also to the spectators,
and might easily be repeated for the sake of the amusement afforded until some
inventive genius turned it into a veritable play with board and men.3
? In this work, Culin systematically calls the game-boards used in American Indian race-games
‘counting-board or abacus’, even when they are marked out on the ground by a circle of stones
or a square composed of an outline of holes in the ground.
2 p. 206. 3 p. 196.
 ORIGIN OF BOARD-GAMES 235
The evidence of existing games and surviving ancient game-materials also
gives little help beyond suggesting that board-games originated in different
ways in various parts of the world. In the Old World, all the leading board-
games, chess, draughts, wei-k’i, fox and geese, and the game of goose, were
invented solely for purposes of recreation. They are essentially pastimes. On
the other hand, a few minor games suggest a different origin. Some are
associated with festivals of various kinds. Thus, the Kanakura tribe of
northern Nigeria, plays canonical games at their annual festival at the end of
the first millet harvest, September or October.’ Chaturaji, the Indian four-
handed dice-chess, was played in the eighteenth century at the festival of the
New Moon, when worshippers kept vigil all through the night (HC. 74).
Women in Ceylon play olinda (mancala) at the New Year (p. 170). More
often, board-games are played during wakes and funeral ceremonies, galat-
jang (7.4.14) in Celebes, mancala by Negroes in Dutch Guiana, which
Herskovits suggests is a custom brought by slaves from West Africa (p. 159).
Mancala boards form part of the furniture of Egbo houses in Calabar (p. 189),
and the mancala board which K. C. Murray saw at the shrine of Odudua in
the village of Iloru, Abeokuta Pr., Nigeria may have been used for divination,
as the Tuaregs of the Sahara play alkarhat (3.6.5) for this purpose, and the
priests in Madagascar played fanorona (4.6.1) during the siege of the capital
by the French for guidance and success in its defence. Since, however,
fanorona was only invented about 1680, this can only carry weight if it
perpetuated a similar use of the parent game, alquerque (4.2.1). There is no
evidence that this game, which is widely played in Asia, was ever played
except as a pastime. I have already mentioned in Chapter 2 the suggestions
that the el-Mahasna board (Egypt) belonged to a magician and that the board
for dogs and jackals (Ur) may have had some religious significance: but there
is no evidence that any Egyptian or Sumerian board-game was played except
as a pastime.
When we turn to the New World, there is more evidence that the race-
games of the American Indian tribes had a religious aspect. The Zufii Indians
play seven games, only one of them a board-game, which are called ‘sacred’,
and with small variations the sacred board-game is played by most of the
Indian tribes. Sho‘liwe (6.7.28) was formerly played by the priests of the bow
at the semi-annual festival of the Twain gods of war, and ta‘sho‘liwe (6.7.20),
the secular form of this game, is still played (Culin, d, 782) as ‘a tribal divina-
tion, a forecast for war or peace, for prosperity or adversity’. The Tewa tribe
in New Mexico plays caseheapana (6.7.27) all the night of 3 November, ‘the
day of the dead’.
Board-games are a species of game and games are merely organized play.

' C. K. Meek, Tribal Studies in Northern Nigeria, 1931, ii. 314. Three games are played on a
board represented by holes made in sand, with four pieces (red and white seeds). Seven elders of
the town, five representing the chief and two, the commoners, take part. Meek calls the game
backgammon, but it may be mancala, as other observers have called mancala ‘backgammon’.
236 THE DISTRIBUTION AND [CH.9
We must, therefore, begin our search for their origins, with an examination
of play, what play is, and what are the circumstances which permit of play.
Play is not confined to man since animals and birds also play. In its simplest
form as played by young children, puppies, and other young animals, play
is mere rough-and-tumble, but some animals and birds have already deve-
loped higher forms of play, such as the ‘dances’ of birds, and children begin
early to imitate the actions of adult men and women. In all these cases, life
expresses itself in action, in doing something, and all play is a form of action.
Adult men play games that are regulated by definite rules which have to be
observed or they cease to be games. There are many kinds of games, but the
most important are the athletic games in which the players exercise their
physical powers which I have briefly dealt with above, and the sedentary
games, board-games, card-games, dominoes, &c., in which the players exer-
cise their mental powers. J. Huizinga (Homo Ludens, 1949, p. 13) defines the
leading characteristics of adult play as being either a contest for something,
or a representation of something. Both aspects are combined in board-games
which are both a representation of human activities and a contest for the
prize of victory.
Groos and Huizinga both start their analysis of play from the assumption
that play is a distinct and highly important factor in man’s life and doings,
and proceed to look for those distinctive features which are typical of play
and differentiate it from other forms of doing. But play refuses to be isolated
in this way. None of the features which they regard as peculiar to play is so;
all are to be found also in work or what Huizinga calls ‘ordinary life’. The
difference between play and work cannot be explained simply on physio-
logical or biological or psychological grounds; it is essentially and only one
of purpose. Every human activity can be work in some circumstances and
play in others. In the beginnings of a civilization, the purpose of a game may
be clearly marked and differentiated from the purpose of work; against the
complicated background of a highly developed civilization, the purpose
becomes blurred; games are no longer played as a religious or magical rite
or only as a pastime or means of recreation, but they may be played as a
livelihood or a business. The professional chessplayer plays chess and the pro-
fessional cricketer plays cricket to earn a living; association football as played
by the Leagues is now a business. So far have we travelled from the original
idea that a game is a pastime or recreation that we may speak of a scholar as
finding his recreation in the study of Ecclesiastical History! If we want to find
out how play and games originated, we must go back at least to the begin-
nings of a civilization.
Play requires no special powers; all that primitive man needs for play is
the opportunity. So long as he is fully occupied in the struggle to keep himself
and his family alive, he has no time for play. The opportunity can only occur
when he has time to spare in his waking hours. Children of tender years
have no responsibilities and have all the day for play, but this does not mean
 ORIGIN OF BOARD-GAMES 237
that adult play evolved from the play of children. The adult does not imitate
the activities of the child.
The circumstances which gave primitive man the leisure for play are the
same as those that are necessary for the growth of a civilization. He must
live in a place where the soil is such that he can make his livelihood easily,
where the climate is neither too hot nor too cold so that work is possible all
the year round, and the satisfaction of his prime needs—food, clothing,
housing, and fuel—only occupy part of the day, and when some social
integration has been achieved for security and for emotional and intellectual
satisfaction, including marital ties, religion, and relaxation. All these condi-
tions are most easily satisfied in the Sub-tropical Zones of the world, and to
the north of the Equator since the greater land-masses lie there, and we have
already seen that it is in these Zones that we find the oldest evidence for the
practice of board-games and that the oldest civilizations of which we have
cognizance arose. It is this coincidence which justifies Huizinga’s claim! that
“games are a cultural phenomenon’.
In the heat of the day when work in the open air is too arduous, or when
the day’s work is over and the daily needs of his family are met, man’s innate
urge to be doing something still impels him to action, if only to the handling
of objects at hand, whether natural like pebbles, or some of his household goods
of his own making, at first aimlessly, but as soon as his attention is held, to
explore their capabilities for new uses. I suggest that it was in this way that
the habit of using objects at hand as playthings, and so as materials for games
arose. Among these objects may have been the lined boards that are still Vv
used in Ceylon as charms and defences against evil spirits and have provided
the boards for games of alinement. We gain some knowledge of the posses-
sions of primitive man from burial mounds and the sites of lake-villages, and a
among the objects discovered from these monuments occur fashioned dice,
both long and cubical, which have been found in the Scandinavian countries,
England,? Ireland, and France unaccompanied by other game-materials,
which suggest that, whatever their original use may have been, they soon
began to serve secular purposes. Liiders (p. 1) shows that chariot-races and
gambling were the leading amusements in Vedic India, and one of the few
secular poems in the Rig-Veda is the lament of a gambler who is unable to
tear himself away from the dice or lots, although he is fully conscious of the
ruin he is bringing on himself and his home (Ragozin, Vedic India, 95).
Tacitus (Germ. 24) also tells how the German tribes in his day were addicted
to gambling. All this points to the conclusion that fashioned lots formed part
of the cherished possessions of man in the early stages of civilization, and that
the handling of these possessions in leisure hours resulted in their use for
games.

2 Long dice have been found in the lake-village at Meare, Somerset (The Times, 3 June 1914
and 8 Nov. 1943).
 238 DISTRIBUTION AND ORIGIN OF BOARD-GAMES [CH.9
Man is naturally gregarious, and learnt early to combine with his fellows
for attack and defence against enemies, for the hunt for food, and for the
cultivation of the soil. So much of his life must have been spent in common
action with his neighbours, that it must have been a natural consequence for
him to share his leisure hours with his fellows and to find both recreation
and amusement in playing with them. In this way, I believe that most
_ board-games originated.
That this suggestion is not far-fetched may be seen from the existence of
string-games. String is a material on which primitive culture depended
greatly, particularly among those peoples who lived by the sea or in islands,
fished for food, and used boats for communication, but also among inland
people. In addition to the ordinary uses made of string, it is used widely for
games of cat’s cradles in which the string is arranged on the fingers and
manipulated to produce figures of animals and other kinds of objects by all
the peoples of Oceania and in Australia, Africa, and both Americas. Mrs.
Rishbeth (Kathleen Haddon) in her Artists in String (1930, p. 151) gives an
analysis of the objects represented in string-games which includes some
1,600 figures, of which one-third are animals and over a hundred are celestial
(the sun, moon, and groups of stars). Unlike board-games, string-games are
independent of climate: they are the only pastimes of the Eskimos. They are
for the most part played by a single person, thus requiring less social integra-
tion than board-games, and they tend to fade away before the higher lights
of civilization. Although the earliest mention of string-games that I know
only occurs in Capt. Bligh, Log of the Voyage of the Bounty, 1787-90, in
which he notes seeing natives of Tahiti playing games with string, the games
must be much older from their wide distribution. Their invention is clearly
the result of handling and playing with string, and supports my view that
other games originated in the same way.
LABUEB SHOWING THE DiS TRIDUEION
OF BOARD-GAMES
240 THE DISTRIBUTION AND

ee
sausddypyq

Se
visauopuy AO

|e es
Tes -VEOa NE)

mM a4

Se
Q mM a4

(Sane eae oe
(a) mM a4

fa

Olm

eee

SHWVO-duvVOd
Gee ees
(a)
uojkan |
< Q

ee ees
pIpuy ysnog me A

PIPUT YI4ON] <a A

AES

G
or

NOILNAIMLSI
ee Oe

AO
puds
ee

Sqvspy <Q Q
EE

DUIYD quajuy <9 Q

Ne

vIpUuy Juaup

auioy iapuy Bomee) isa)

ee

a0aas5) JuaIIUy <a Q AQ

ee

auljsaod AQ
eee
eS

iddBq quavuy <a Q isa} ae;

:
ee
soure3-piedoaq

:
xoJ puke asaay)
xo, pur 98905)

soure3-ro3ry,
soure3-junzy

*
JoTTeEUrs SPIOP]

I93IVeT S[OIOJY

soure3-uonoulolg
eee

9800H
QUINT SOTOH]

WIOpopy
*

*
sparoyy

JuotUy

enbionbry epedeqysy
syysneig Jo
suey
esis
e100
* BEL eigen
ssayD qe indneyd
$

‘eyoueg
“o29 ||
essed
pelea,
|
||
ee
epeouryy
Jy
IX*:2Ww
P[eoue
III;5 breoueyy
AI
chiles
:
"
5356
1d

Aupusay
uIsvg osuoD

vatduvsuv J.

pury|
puojooy|
OUINN STOR]

pduay|

BELOrn |<
piapuipuvIg |
adosnq isvq-yjnog |
vissnyy |
vuufy wines|

<q
wopns |
ppuvs) |
“222 ‘029040 |
Sa0s5aNT wosuouy |

IoTTews SPIO]

punypouos| <i

sours] <0
piuissdqy | —Q

won| <0
$ SfaIayAy
*

svasvsvpoyy | ~Q
supipuy wosuowy | a)

pasfy 15044 |=O

Jo3reT S[erspy
*

a
a
e1019g

upds | se) a |
3 |
BeL |
puo5in| <Mm0Q |
A |

|mu
A |me.
RSIS

)
O
ORIGIN

anbionbry ° s |) {Be || St
syysneiq
C c . I I

—
=

_—
_—
pa
ssaya ne co feteeEg)
OF

TOPPA
* : ; .

soy ile
HSPs)

BB lim)
qeL . e : ; I al a
soured-junzy
* a Wi:
JuaIUy
xoj puke 98995)
* NININI
N NININININ
ulopoypy
xoj puke 28995) OTTOwc@, O
soue3-piedoaT
soure3-1931[,
eR
7 : ; % i) Rel |p Ret |) tel |] Ral
epedeiysy : :
BOARD-GAMES

-indneyD
Q : : L
AT
“eyoueg
"229 . . .
suey
Jo 98005
: Ke
| AIS IN WN
souied-uonowo
: ig
eryryeg
- : ; |- x
ejesueyy
I] ; ; ; AS | AS | AK | AS | AX
eeoueyy“TI ; ; Zz
N

Pfeoueyy
AI
N

Gs a7| 7 | Wad
|>N&

NOLLNGIML
JO SIG SHWVO-duvOd
241
 BOOKS AND ARTICLES CONSULTED

Aszott, G. F., Macedonian Folklore, Cambridge 1903.

- Anprée, R., Ethnographische Parallelen, Neue Folge, Leipzig 1899.
Annual ofBritish School at Athens, vii (1900-1).
Ault, N., Life in Ancient Britain, London 1920.
Austin, R. G., (a) ‘Roman Board Games’, in Greece and Rome, iv
(1934-5).
—— (b) ‘Zeno’s Game of Tablé’, in Journal ofHellenic Studies, liv (1934).
— (c) ‘A Roman Game-board’, in Archaeologia Cambrensis (1939).
—— (d) ‘Greek Board-games’, in Antiquity (1940).
Ave or, R., (a) ‘Le Jeu des godets’, in Bulletin de la Société d’Anthropologie de
Paris, Fifth Ser., vii (1906).
—— (b) ‘L’Ouri’, ibid., Fifth Ser., ix (1908).
_~ Baspen, G. T., Among the Ibos of Nigeria, London 1921.
BeEcQ DE Fouqurirss, L., Les Jeux des Anciens, Paris 1869.
_+ BELGRAVE, C. D., Siva, London 1923.
_— Bennett, C. H., Games of Skill and Conjuring, London, New York 186r.
BennelT, G. T., “The Game of Wari’, in Rattray, R. S., Religion and Art in
Ashanti (Oxford 1927), pp. 382-90.
_~ Bent, J. T., The Sacred City of the Ethiopians, London 1893.
Best, Erspon, The Maori, iv (1924).
- Boyprn, E. C., The New Backgammon, London 1930.
Boy’s Own Book, London 1868.
— Branb, J., Observations on Popular Antiquities, Newcastle 1777.
BrAUNHOLTZ, H. J., (a) “The Game of Mweso in Uganda’, in Man, xxxi
(1931).
—— (b) ‘Note’, ibid. xxxi (1931).
Briercey, H., “Ancient Game of Merrils’, in Yorkshire Weekly Post, 3 July
1915.
British Museum Quarterly, ii (1927).
Brunet Y BELtet, J., El Ajedrez, Barcelona 1891.
— Buttock, C., The Mashona, Cape Town 1927.
_— Burton, R. F., First Footprints in East Africa, London 1856.
—~Biurtixorsr, J., ‘Einiges iiber die Eingebornen von Liberia’, in Internatio-
nale Archiv ftir Ethnographie (1882).
— CAMPBELL, D. M., Java Past and Present, London 1915.
_ CarDINALL, A. W., In Ashanti and Beyond, London 1927.
— CARNARVON, Eart of, and Carter, H., Five Years’ Excavation at Thebes,
Oxford 1912.
Carrera, P., I] Gioco degli Scacchi, Militello 1617.
Cassell’s Book of Sports and Pastimes, London.
Cuo Yo, Japanese Chess, Chicago 1905.
 BOOKS AND ARTICLES CONSULTED 243
Coun, M., ‘Jeux Abyssin’, in Journal Asiatique, toth Ser., xviii (1911).
Cormac’s Glossary, edited by W. Stokes, London 1862.
_— Covarrusias, Tesoro de la Lengua Castellana, 1611.
CRUICKSHANK, J. G., “British Guiana Games’, in Man, xxix (1929).
Cun, Stewart, (a) Korean Games, Philadelphia 189s.
—— (b) Chinese Games with Dice and Dominoes, Washington 1895.
(c) Mancala, Washington 1896.
—— (d) Chess and Playing Cards, Washington 1898.
(e) ‘Hawaiian Games’, in American Anthropologist, New Ser., i (1899).
—— (f) Philippine Games, New York 1900.
——(g) Games of the North American Indians, Washington 1907.
DANDEMON, ‘Jeux malgaches’, in Bulletin de ’Académie Malgache, vi (1909).
DartTneELL, G. E., ‘Sixpenny Madell’, in Notes and Queries, Eighth Ser., xii
(1897).
Datta, JOTINDRA Mouan, ‘Sedentary Games of Lower Bengal’, in Journal
of the Asiatic Society of Bengal, New Ser., xxix (1933).
Davies, R., “Some Arab Games and Riddles’, in Sudan Notes and Records, viii
(1925).
—Divertissemens Innocens, La Haye 1696.
_~ Doveuty, C. M., Travels in Arabia Deserta, Cambridge 1888.
_— DRAKE-BROCKMAN, R. E., British Somaliland, London 1912.
Drwerc, J. H., “The Game of Choro or Pereaiini’, in Man, xxvii (1927).
~~ Dusors, Monographie des Betsileo, 1938.
— Du Cuattu, P. B., The Viking Age, London 1899.
Dural, Mrs. H. G., ‘Pallanguli’, in Man, xxvii (1928).
Dussex, O. T., “Notes on Malay Games’, in Journal of the Royal Asiatic
Society, Straits, xxx (1938). £ wv © Ala) Der dyy
EcuarevaA, J. U., Benin Games, Benin 1949.
— Emppen, E. van, Verhandeling over het Damspel, Amsterdam 1787.
Ems ery, E. S., “The Turkana of Kolosa District’, in Journal ofthe Royal Anthro-
pological Institution, lvii (1927).
Emery, W. B., Nubian Treasure, London 1948.
— Encyclopedia of Sports, Games and Pastimes, London 1935.
— Encyclopédie Méthodique. Dictionnaire des Jeux, Paris 1792.
_— EnGELHARDS, H. E. D., ‘Mededeelingen over het Eiland Saleijer’, in Bijdragen
tot de Taal-, Land-, en Volkenkunde van Nederlandsch Indié, Fourth Ser.,
viii (1884).
FALKENER, E., Games Ancient and Oriental, London 1892.
FELKIN, R. W., ‘Notes on the For Tribe’, in Proceedings of the Royal
Society ofEdinburgh, 1885.
Fiske, D. W., Chess in Iceland, Florence 1905.
_— Frercuer, H., Hausa Sayings and Folklore, London 1942.
Gapp, C. J., ‘An Egyptian Game in Assyria’, in Iraq, i (1934).
_— GaLtt, J., Letters from the Levant, London 1813.
 244 BOOKS AND ARTICLES CONSULTED
_» GirALDus CAMBRENSIS, Gemma Ecclesiastica, Rolls Ser., London.
Gopparp, A. R., ‘Nine Men’s Morris’, in Viking Club’s Saga-book, Coventry
1901.
Gomme, Auice, Games Traditional, London 1894.
— Grand Trictrac, le, Paris 1756.
GrIAULE, M., (a) Jeux et divertissements abyssins, Paris 1935.
—— (b) Jeux dogons, Paris 1938.
—Groos, K., The Play of Man, translated by E. L. Baldwin, London rg9or.
Groves, Capt. E. T. N., ‘Customs of the Acholi’, in Sudan Notes and Records,
ii (1919).
Gupta, Hem Cuanpra Das, (a) ‘Bara guti (Punjab) and Similar Games in
East Bengal’, in Calcutta Review, vi (1923).
—— (b) ‘Hole and Pebble Games in Assam’, in Journal of the Asiatic Society
of Bengal, New Ser., xix (1923).
(c) ‘Draughts &c. in Central Provinces’, ibid. xx (1924).
——(d) ‘Games from the Punjab, Central Provinces and Orissa, ibid.
xxii (1926).
Happon, A. C., Study of Man, London 1896.
— Happon, KatTHLeen (Mrs. Rishbeth), Artists in String, London 1930.
— Haun, A. von, Buch der Spiele, 3e Aufl. Leipzig 1900.
Heuer, M., ‘Notes on the Malay Game Jongkak’, in Journal of the Royal
Asiatic Society, Straits, xlix (1907).
— Herxtots, G. A., Qanoon-e-Islam, London 1832.
Herskovits, M.J., (a) “Adjiboto’, in Man, xxix (1929).
—— (b) ‘Wari in the New World’, in Journal of the Royal Anthropological
Institute, lxii (1932).
_—Hmnty, K., (a) ‘Anmerkungen in Beziehung auf das Schach und andere
Brettspiele’, in Zeitschrift der Deutschen Morgenlandischen Gesellschaft,
a
xxvii (1873).
—— (b) ‘Morgenlandisch oder Abendlindisch?’, ibid. xliii (1889).
~— —— (c) ‘Einige Worte iiber das persische Brettspiel Nard’, ibid. xxxiii (1879).
Hosiery, C. W., (a) Ethnology of A-Kamba and other East African Tribes,
Cambridge 1910.
a
—— (b) Eastern Uganda: an Ethnological Survey, London 1902.
Hopson, T. C., The Naga Tribes of Manipur, London 1911.
— Hormg, R., (a) Academy of Armory (1681), Roxburghe Club (1905).
——(b) Academy of Armory, Chester 1686.
Hora, SUNDAR Lat, ‘Sedentary Games of India’, in Journal of the Asiatic
Society of Bengal, New Ser., xxix (1933).
— Hoytg, E., Backgammon, London 1745.
Huizinea, J., Homo Ludens, translated by R. F. C. Hull. London 1949.
Humpurigs, E. pe M., “Notes on Pachesi and Similar Games as played in
Karwi Subdivision, United Provinces’, in Journal of the Asiatic Society of
Bengal, New Ser., ii (1906).
 BOOKS AND ARTICLES CONSULTED 245
Humpineg, K. P., “Shayo’, in Nigeria (1938).
Hutton, J. R., (a) The Angami Nagas, London 1921.
—— (b) The Sema Nagas, London 1921.
Hype, T., De Ludis Orientalibus Libri Duo, Oxford 1694.
— IncraMs, W.H., Zanzibar, London 1921.
IstorE, Bishop, Origines in Opera Omnia, Rome 1797-1803.
— Islenzkar Gatur, iv, Kaupmannahéfn 1892.
_— Jonson, R., The Golden Trade, 1623.
_— Jouns, C. N., Excavations at Pilgrim’s Castle, Atlit, London 1933.
_ Jounston, Sir H., (a) George Grenfell and the Congo, London 1908.
_/ —— (b) ‘Survey of the Ethnology of Africa’, in Journal ofthe Royal Anthropo-
logical Institute, xliii (1913).
— Journal of theManx Museum, Douglas, ii (1934).
_~ Junop, R. A., Life of a South African Tribe, second edition, London 1927.
_- Kaupern, W., Ethnographical Studies in Celebes, ’s~Gravenhage 1929.
Kennepy, W. P. M., Elizabethan Episcopal Administration, London 1924.
- Kircuine, A. L., The Backwaters of the Nile, London 1912.
_— Koutmann, P., (a) Der Nordwesten unserer Ostafricanischen Kolonie, Berlin
1898.
_~ —— (b) The Victoria Nyanza, London 1899.
_~ Komarov, P., ‘Turkestanskiya Shashki’, in Shakhmatnoe Obozrenie, iii (1893).
_— Korscue1t, O., ‘Das Go-Spiel’, in Mittheilungen der deutschen Gesellschaft fur
Natur- und Volkerkunde Ostasiens (1880-1).
_— Kraitcuik, M., Mathematical Recreations, London 1948.
_~ Lata Raja Basu, Mo’allim ul Shatranj, Delhi 1901.
Lang, E. W., Modern Egyptians, Minerva Library, second edition, London
1890.
_—— Larkin, Major P. M., “Account of the Zande’, in Sudan Notes and Records, ix
(1926).
_- Leakey, L. S. B., White African, London 1937.
Lewis, F. R., ‘Gwerin ffristial a Thawlbwrdd’, in Cymmrodorion (1941).
Limprey, G. H., Ancient History of the Game of Draughts, London 1913.
_— Linpstom, G., The Akamba in British East Africa, Uppsala 1920.
Linpg, A. VAN DkR, (a) Geschichte und Litteratur des Schachspiels, Berlin 1874. 9chess
—— (b) Quellenstudien, Berlin 188t.
Linton, R., The Tanala Hill Tribe of Madagascar, Chicago 1933.
Low, Capt. J., ‘On Siamese Literature, Part II’, in Asiatic Researches, xx
(1836).
_— Lowrz, R. M., History ofEthnological Theory, London 1930.
ao _— Lucas, T., Memoirs of the Lives of the most Famous Gamesters, London 1714. ;
Lipers, H., Das Wiirfelspiel im alten Indien, Berlin 1907. Di cé
Lupovist, L., ‘Sports and Games of the Singhalese’, in Journal of the Royal
Asiatic Society, Ceylon (1873).
_- Lumuo1irz, C., Through Central Borneo, London 1920.
 246 BOOKS AND ARTICLES CONSULTED
LuscHan, F. von, (a) ‘Bericht iiber eine Reise inSiid-Afrika’, in Zeitung fiir
Ethnologie, xxxviii (1906).
— (b) Die Altertiimer von Benin, Berlin and Leipzig 1918.
— (c) ‘Mankala-Spiel’, in Mittheilungen der Wiener Anthropologischen Gesell-
schaft, xlviii (1919).
Maca ister, R. A. S., (a) History ofCivilization in Palestine, Cambridge 1921.
—— (b) A Century ofExcavation in Palestine, London 1925.
(c) Archaeology of Ireland, London 1928.
—— Macxenzig, D. R., The Spirit-ridden Konde, London 1925.
~ Maison des Jeux académiques, Paris 1668.
_— Mattst, P., Le Ieu des Dames, Paris 1668.
—— Manoury, M., Le Jeu de Dames a la Polonaise, Paris 1787.
Mar, G., (a) ‘Somali Games’, in Journal ofthe Royal Anthropological Institute,
lxi (1931). eee:
—— (b) ‘An Ancestor of the Game of Ludo’, in Man,(1xii)(1940). YY
_-~ Mek, C. K., (a) Tribal Studies in Northern Nigeria, London 1931.
—— (b) A Sudanese Kingdom, London 1931.
— MerkKer, O., Die Masai, Berlin 1904.
— Meyer, H., Die Barundi, Leipzig 1916.
— Microp, F. W. H., View of Sierra Leone, London 1915.
— Mockter, Capt. A. FerryMan, Up the Niger, London 1892.
Monckton, O. P., Pastimes in Times Past, London 1913.
Monop, Th., ‘Sur quelques jeux africains’, in Notes africaines (Jan. 1950).
_— Montettus, O., Kulturgeschichte Schwedens, Leipzig 1906.
Montcomery, W., “The Malagasy Game of Fanorona’, in Antananarivo
Annual, iii (1896).
Mosso, A., Palaces of Crete, London 1907.
_— Moutipars, T. pe, Grande Encyclopédie des Jeux, Paris 1840.
Murray, A. S., Smit, A. H., and Watters, H. B., Excavations in Cyprus,
London 1900.
Murray, H. J. R., A History of Chess, Oxford 1913.
Newserry, R. J., “Games and Pastimes of Southern Nigeria’, in Nigerian
Field, viii (1939).
_— Nriepunr, K., Reisebeschreibung von Arabien und anderen umliegenden Landern,
Copenhagen 1774-8.
_—— OrDE Browne, G. St. J. J., The vanishing tribes of Kenya, London 1925.
Overseck, H., ‘New Notes on the Game of Chongkak’, in Journal of the
Royal Asiatic Society, Straits, xviii (1915).
_~ OVERBERGH, M. C., Mangbetu, Bruxelles 1909.
— Patras, P. S., Sammlungen historischer Nachrichten iiber die Mongolischen
Volkerschaften, St. Petersburg 1776.
Parker, H., Ancient Ceylon, London 1909.
_— Parry, N., The Lakhars, London 1932.
_+ PARTRIDGE, C., Cross River Natives, London 1905.
 BOOKS AND ARTICLES CONSULTED 247
PENDLEBURY, J. D. S., Archaeology of Crete, London 1939.
Percy, R. and S., Percy Anecdotes, Chandos Library, London.
Petrorr, A. D., Rukovodstvo, St. Petersburg 1827.
Puitscuke, K., ‘Kurze Mittheilung iiber zwei Malayische Spiele’, in Inter-
nationales Archiv fur Ethnographie, iii (1893).
_ PLowpsn, W. C., Travels in Abyssinia, London 1868.
Powe t-Corton, P. H. G., ‘A Mancala Board called Songo’, in Man, xxxi
(1931). ;
— QuerciTano, D., L’Egide de Pallas, Paris 1727.
— RaBELals, F., Liure Premier (1535), Paris 1835.
— Raretes, Sir T. SraMrorD, History ofJava, London 1817.
Ricuter, W., Spiele der Griechen und Romer, Leipzig 1887.
~ Rosinson, H. O., ‘Dum Blas’ in Cheltenham Examiner, 19, 28 Nov. 1903.
~ - Rosinson, J. AnmiTacGE, Time of St. Dunstan, Oxford 1923.
~~ Ropp, F. R., People of the Veil, London 1926.
~— Roy, C. F., Unconquered Abyssinia, London 1923.
~— SALISBURY, JOHN OF, Policraticus.
SANDERSON, M. G., “Native Games of Central Africa’, in Journal of the Royal
Anthropological Institute, xliii (1913). 1423
_- SaraT Cuanpra Das, Journey to Lhasa, edited by W. W. Rockhill, London
1904.
SAVENKOV, E. V., K Voprusu op Evolutsie Shakhmatnoi Igry, Moscow 1905.
——~ ScHAPURA, I., Bantu-speaking Tribes of South Africa, London 1937.
— SCHUEMANN, H., Troy, translated by Dr. P. Smith, London 1875.
— ScHuttze, L., Aus Namaland und Kalahari, Jena 1907.
— ScHWEINFURTH, G. H., Artes Africanae, Leipzig 1873.
-—— SELENUS, G., Das Schach- oder Kénig-Spiel, Leipzig 1616.
_— Seymour, R., Court Gamester, London 1754.
SHACKELL, R. S., (a) ‘Mweso’, in Uganda Journal, ii (1934).
—— (b) ‘More about Mweso’, ibid. iii (1935).
_ SHAKESPEAR, J., Lushei Kuki Clans, London 1912.
SHEPPARD, T., ‘A Mweso Board from Mombasa’, in Man, xxxi (1931).
_— Suirwoon, J., De Ludo Arithmomachia (MS. Rome, Casanatense 791).
_—— Suway Yok (Sir G. G. Scott), The Burman, London 1910.
S1eBER, D. J., ‘Das Leben der Kinder in Nsungli-Stamm’, in Afrika, xi (1938).
SkEaT, W. W., Malay Magic, London 1900.
Smitu, A., The Game ofGo, New York 1908.
_~ Smitu, E. W. and Dat, A. M., The Ila-speaking Peoples ofNorthern Rhodesia,
London 1920.
— Smitu, H. A., Guide to the A.S. Antiquities, British Museum, London 1920.
— $itH, J. H., Lachesis Lapponica, London 1811.
— Snouck-Hureronje, C., The Achehnese, translated by A. W. S. O’Sullivan.
Leyden & London 1906.
_— Srayt, H. A., The BaVenda, Oxford 1931.
 248 BOOKS AND ARTICLES CONSULTED
Srrutt, J., Sports and Pastimes of the People ofEngland, edited by W. Hone.
1838 (London 1876).
_— Sykes, Etta C., Persia and its People, London 1910.
——Tatsot, P. A., The Peoples of Southern Nigeria, London 1926.
—- TESSMANN, G., Die Pangwe, Berlin 1912.
Torpay, E. and Joyce, T. A., ‘Les Bushongo’, in Annales du Musée du Congo
Belge, sér. iii, tome ii (1911).
—TREMEARNE, A. J. N., Hausa Superstitions and Customs, London 1913.
_— Tressan, L., Das Schachspiel, seine Gattungen und Abarten, Quedlinburg and
Leipzig 1840.
—— Treves, Sir G., Uganda for a Holiday, London 1910.
_— Twiss, R., Miscellanies, London 1805.
_— Tytor, Sir E. B., Anthropology, London 188.
VANOVERBERGH, M., ‘Iloko Games’, in Anthropos, xxii.
-— Vetula, Liber de, Wolfenbiittel 1661.
Waytanpb, E. J., (a) ‘Preliminary Studies of the Tribes of Karamoja’, in
Journal of the Royal Anthropological Institute, lxi (1931).
(6) “Mweso’, in Uganda Journal, iv (1936).
— Weeks, J. H., Among Congo Cannibals, London 1913.
Werner, O., Natives ofBritish Central Africa, London 1906.
WIEDERMANN, ‘Das Brettspiel bei den alten Aegyptern’, in Actes du 10°
Congrés d’ Orientalistes, 1897.
Wixinson, R. J., Papers on Malay Subjects, Kuala Lumpur 1915.
_— Wison, Tom, Illustrerad Spelbok, Stockholm 1888.
Wintock, H. R., Excavations at Deir el-Bahri, 1925-6, New York 1942.
Wool tty, Sir L., Ur, the First Phases, London 1946.
_~ Woop-Manrtin, W. G., Lake Dwellings of Ireland, Dublin 1886.
 INDEX, OF “GAWNEES
Abanga (Niam-niam), 7.7.1. Awithlaknakwe (NAI.), 4.1.14.
Acey-deucy (U.S.A), 6.2.48. Awithlaknan(n)ai (NAI.), 3.5.29; 4.2.17,
Achi (Gold Coast), 3.4.4. 43-
Adangang (Celebes), 6.1.6. Awl Game (NAI.), 6.7.11-35.
Adenagank (NAI.), 4.3.2. Awli on-nam ot-jin (Borneo), 7.4.13.
Adi (Togoland, Gold Coast), 7.5.28, 36. Ayo (Dahomey, Nigeria), 7.5.16, 32.
Adji (Gold Coast, Dutch Guiana), 7.5.28, Ayoayo (Nigeria), 7.5.17.
36; 7.9.5. Azigo (Nigeria), 7.5.46, 47, 50.
Adjiboto (Dutch Guiana), 7.9.4.
Adji-pre (Dutch Guiana), 7.9.5. Babylonian (France, Holland), 4.3.7.
Adjito (Dahomey), 7.5.10. Backgammon (England), 6.2.12; French
Adzua (Nigeria), 7.5.40. or Double (England), 6.2.47; (German
Aghadaghada (Nigeria), 7.5.7. or Russian), 6.2.44.
Agsinnoninka (Philippines), 7.4.17. Bagh bandi (Lower Bengal), 5.6.9.
Aiyawatstani (NAL), 4.2.16. Bagh bukri (Deccan), 5.6.15.
Akidada (Gold Coast, Nigeria), 3.1.11; Bagh guti (India, C.P.), 5.6.10.
3.4.43 3.5.26. Bahendagabo (Tanganyika), 8.3.5.
Akileyisit (Uganda), 7.6.17. Bajangkak (Sumatra), 7.4.7.
Ako-okwe (Nigeria), 7.5.56. Bakkis (Somaliland), 6.4.13.
Akuyo (NAI.), 2.7.3. Bangale (India C.P.), 4.2.22.
Alea (Rome), 2.8.3; 6.1.1. Banya (Sudan), 8.5.2.
Alkarhat (Sahara), 3.6.5. Bao (Zanzibar), 8.3.11.
Alquerque (Anc. Egypt), 2.1.10; (Europe, Bao kiarabu (Zanzibar), 8.2.1-2.
Asia), 4.2. Bao kiswahili (Zanzibar), 8.3.11; (Congo),
Alquerque de doze (Spain), 4.2.1; de Txsolte
nueve, 3.5.1; de tres, 3.3.2. Baqira (Arabs), 7.1.3.
Ambiddo (A-Madi), 7.7.5. Bara-guti (Punjab), 4.2.7.
Anglicorum, ludus (England), 6.2.25. Bara-guti pait pait (Bengal), 3.5.17.
Apit, apit-sodok (Malaya), 4.1.5. Barail (Italy), 6.2.13-15.
‘aqil, lib al- (Egypt), 7.1.7; (Damascus), Barignin (France), page 119.
Gfaibetoxs Barjis (Arabs, Palestine), 6.4.12.
Arb‘ata ‘ashara (Arabs), 8.1.1. Bau (Zanzibar), 8.3.11.
Aringari (Fors), 7.5.8. Baz (India, U.P.), 4.2.22.
Arneb (Sudan), 7.5.5. Bazi qamash (Persia), 4.10.1.
Asalto (Europe), 5.3.7. Beatta (Arabs), 7.1.11.
Ashere (Nigeria), page 14 n. Belagerungs-spiel (Germany), 5.3.7.
Ashta-kashte (Bengal), 6.3.3. Beligringsspel (Sweden), 5.3.7.
Ashtapada (India), 6.3.1. Beleirings-spil (Iceland), §.3.7.
Asi keliya (Ceylon), 6.4.15. Belejringspel (Denmark), 5.3.7.
Assaut (France), 5.3.7. Bethelas (Italy), 6.2.49.
AO elta stelpur (Iceland), 6.2.8. Bhagchal (Assam, Sikkim), 5.6.22.
Athara guti (India, U-.P.), 4.2.22. Bidaria (NAI.), 3.3.20; 3.5.28.
Athara gutiala teoria (India, C.P.), 4.2.22. Bonzo (Congo), 7.8.2.
Ati (Gold Coast), 3.4.4. Boolik (NAI.), 6.7.6.
Atlanbaj (Turkey), 4.4.1. Bosh (Somaliland), 7.6.7.
Awa(ng)du (Dahomey), 4.10.6. Bouge-shodra (Siberia), 5.6.13.
Aware (Ivory Coast), 7.5.15. Briadspel (Svensk), 6.2.39.
Awari (Dutch Guiana), 7.9.3. Brandub(h), Brannaib (Ireland), 2.9.2.
250 INDEX OF GAMES
Brettspiel (Germany), 6.2. Coyote (Mexico), 5.1.4.
Buanfach (Ireland), 2.9.2. Cuwee (NAI.), 6.7.12.
Bub (Somaliland), 4.1.8.
Buf, buffa (Italy, Spain), 6.2.42. Dagboprou (Ivory Coast), 7.5.15.
Buf de Baldriac (Spain), 6.2.45. Dagh (Nigeria), 7.5.20, 34.
Buf regal (Germany), 6.2.50. Dajam (Madras), 6.4.14.
Buffa cortesa (Italy, Spain), 6.2.42. Dakon (Java), 7.4.8.
Bushels (Salop., England), 3.5.3. Dala, Dali (Sudan), 3.6.3.
Dam (Sweden), 4.3.1; (Ceylon), 4.3.9;
Cab e equinal (Spain), 6.2.16-17. (Malaya), 4.2.27.
Camarzo (Italy), 6.2.51. Dam blas (Malaya), 4.3.7.
Canes, Los doze (Spain), 6.2.1. Dam-daman (Java), 4.2.34.
Canis Martius (Italy), 6.2.51. Dama (Italy, Poland, Romania, Egypt,
Carabawg (Somaliland), 3.3.17. Arabs of Palestine, Turkey), 4.3.1, 3, 6.
Carré Chinois (France), 3.3.5. (Also in Czech, Serbian, Slovene.)
Caseheapana (NAI.), 6.7.28. Damapelissa (Finland), 4.3.1.
Cashlan gherta (Ireland), 3.3.7. Damas (Spain, Portugal), 4.3.1, 4.
Castro (Spain), 3.5.1. Dame (Italy, Poland), 4.3.1, 3, 6.
Catchdolt (England), 6.2.4. Damen (Holland), 4.3.1.
Cennchaem Conchobair (Ireland), 2.9.2. Damen jokoa (Basque), 4.3.1.
Cercar la liebre (Spain), 5.1.1. Damenspiel (Germany), 4.3.1.
Chahrachel (Assam, Sikkim), 5.6.22. Damenspil (Sweden), 4.3.1.
Chato (Sumatra), 7.4.5. Dames (France, England, Catalonia), 4.3.1.
Chatush-pada (India), 2.10. Damki (Ukraine), 4.3.1.
Chaupur (India), 6.4.1. Damm (Iceland, Sweden), 4.3.1.
Chauput (India), 6.4.5. Damma (Sahara), 4.2.14, 35.
Chausar (India), 6.4.3. Dammen (Germany, Holland), 4.3.1.
Checkers (England, U.S.A.), 4.3.1. Dammenspiel (Germany), 4.3.1.
Chela (Angola), 7.8.7. Dammes (England), 4.3.1.
Cherrapunji (Assam), 7.2.3. Damo (Provence), 4.3.1.
Chess, 4.5.1. Dams (Scotland), 4.3.1.
Chiana wa bwalo (Nyasaland), 8.4.2. Damspel (Holland, Sweden), 4.3.1.
Chiana wa kunja (Nyasaland), 8.4.1. Damspil (Denmark, Iceland, Norway),
Chikkiri (Ainus, Japan), 3.1.8. AS sls
Chingskoot (NAI.), 6.7.36. Dandaman (Java), 4.2.34.
Chisolo (N. Rhodesia), 8.2.7-9, 18, 20. Dandimannstafl (Iceland), 6.2.52.
Choa, Len (Siam), $.5.1. Danumannstafl (Iceland), 6.2.52.
Choko (Gambia), 4.4.5. Dara-dara (Celebes), 7.4.15.
Chongkak (Malaya), 7.4.4. Daramutu (Ceylon), 7.3.4.
Chonka, chonku (Ceylon), 7.3.5; 7.4.12. Darra (Nigeria), 7.5.41-42.
Choro (Uganda), 8.3.3, Io. Dasapada (India), 2.10; 6.3.1.
Chuba (U.S.A.), 8.2.16. Dash-guti (India, C.P.), 4.2.38.
Chuk sam (China), 3.5.22. Deleb (Somaliland), 4.10.2.
Chuka (Sumatra), 7.4.5. Demala diviyan (Ceylon), 5.5.4.
Chuki (Malaya, Indonesia), 4.9.3. Derra (Nigeria), 7.5.41-42.
Chuncajon (Philippines), 7.4.16, 17. Dhola (Maldive I.), 6.4.10.
Comican (Chile), 4.11.1. Diagrammismos (Greece), 2.7.2.
Contrabuf, Contrare puff (Germany), Dili (Nigeria), 3.6.4.
6.2.47. Diviyalliya (Ceylon), 5.6.8.
Copped crown (England, Cumb.), 3.3.6. Diviyam keliya (Ceylon), 5.6.8.
Coquimbert (France), 4.3.10. Doat, Len (Siam), 6.4.18.
Coro (Uganda), 8.3.3, 10. Doblet (Spain), 6.2.2.
Corsi crown (Scotland, Galloway), 3.3.6. Do-chu-sugoroku (Japan), 6.3.8.
 INDEX OF GAMES 251
Dogs and jackals (Egypt, Ur, Assyria, Hit- Frauenspiel (Germany), 6.2.
tites, Palestine))2:1.$ 3922-2 582-3,10 2e4- Fuchs im Hiihnerhof (Germany), 5.3.1.
Do-guti (Punjab), 4.8.1. Fuchs und Ganse (Hiihner) (Germany),
Dominarum, Ludus (Italy), 4.3.3. S30:
Donne (Italy), 4.3.1.
Doppelmiihle (Germany), 3.5.5. Gaasespil (Denmark), 6.5.5.
Doublets (England), 6.2.5. Gabata, Gabatta (Abyssinia), 7.6.1; 8.1.2.
Dra (Sahara), 3.6.4. Gala (Celebes), 4.1.12; 6.3.3.
Draughts, Canadian, 4.3.8. Galatjang (Celebes), 7.4.14.
Draughts, Diagonal (England), 4.3.11. Gamma (Dogons), 7.5.3.
Draughts (England), 4.3.1. Gammon (Scotland), 6.2.12.
Draughts, Losing game, 4.3.10. Gansespiel (Germany), 6.5.4.
Draughts, Polish, 4.3.7. Ganzenspel (Holland), 6.5.5.
Dniesticken (Holland), 3.1.2. Garanguet (France), 6.2.28.
Dris (Arabs), 3.1.5; 3.3.83 3.5.13, 14. Gavalata (S. India),-6.3.7.
Dubbel-qvarn (Sweden), 3.5.8. Gegenpuf (Germany), 6.2.47.
Dugh (Armenia), 3.5.16. Ghashim (Arabs), 7.1.4.
Duodecim scripta (Rome), 2.8.2. Giacchetto (Italy), 6.2.37.
Duqurjin (Turkey), 3.5.15. Gifia (Nigeria), 7.5.61.
Giuthi (Kenya), 7.6.10.
Ebe (A-Bangba), 7.7.6. Go-bang (Europe), 3.7.2, 3.
Edris a jin (Druses), 6.4.17. Gol-skuish (India, C.P.), 4.2.41.
Egara guti (India, C.P.), 4.2.39. Gomok-narabe (Japan), 3.7.1.
Ekwe (Nigeria), 7.5.59. Goose, Game of (England), 6.5.5.
Ellaewala-kanda (Ceylon), 7.3.4. Grammai (Greece), 2.7.2.
Emperador (Spain), 6.2.25. Grasshopper (Europe), 3.8.1.
Emperador, El medio (Spain), 6.2.27. Gué (A-Barambo), 7.7.8.
En dodoi (Kenya), 7.6.14. Guebeta (Abyssinia), 7.6.1.
En geishei (Kenya), 7.6.14. Guvoota (Abyssinia), 7.6.1.
Epelle (Nigeria), 3.3.18. Gwyddbwyll (Wales), 2.9.1; 4.1.2.
Erherhe (Gabon), 7.5.29.
Escaques (Spain), 6.8.1. Hakimi (Damascus), 7.1.8.
Ezu ahia ako okwe (Nigeria), 7.5.56. Halatafl (Iceland), 5.3.1.
Halma (Europe), 3.8.1.
Faille (France); fails (England), 6.2.18. Halisa, halisi (Arabs), 7.1.1.
Falakiya (Arabs, Persia), 6.8.1. Han (Egypt), 2.1.2.
Fallas (Spain), 6.2.18. Hasami-shogi (Japan), 4.1.6.
Fanorona (Madagascar), 4.6.1. Hat diviyan keliya (Ceylon), 5.5.2.
Fanorona telo (Madagascar), 3.3.19. Helga (Libyan desert), 4.4.6.
Farisia (Moors), 4.3.1. Hermanos, Los doze (Spain), 6.2.1.
Fastningsspel (Sweden), $.3.7. Hewakam keliya (Ceylon), 4.2.26.
Ferses (England), 4.3.1. Hnefatafl (Iceland), 4.1.13.
Festungsspiel (Germany), 5.3.7. Hoppity (England), 3.8.1.
Fidchell (Ireland), 2.9.1; 4.1.2. Hue-ta-quee-che-ka (NAI.), 6.7.22.
Fierges (France), 4.3.1. Hujura (Persia), 3.3.9.
Fifanga (Madagascar), 8.5.10. Hus, Huts (SW. Africa), 8.2.4.
Filetto (Italy), 3.3.4; 3.5.3.
Filo (Italy), 3.3.4; 3.5.3. I (Dogons), 7.5.23, 43, 44.
Filo-mulino (Italy), 3.5.3. I pere (Dogons), 7.5.24.
Forcat (Provence); forcé (France), 4.3.1. Iblis, Li‘bat (Sudan), 7.5.2.
Forkoering (Denmark), 6.2.39. Ichwe (Nigeria), 7.5.18.
Fox and geese (Britain), 5.3.1; Double, I-go (Japan), 4.7.1.
Ce pyle Mewes, SS Igori (Nigeria), 7.5.19.
 Ruben 2s

252 INDEX OF GAMES

Ikibuguzo (Tanganyika, Urundi), 8.3.4. Komari (Uganda), 8.1.4.
Ikisoro (Tanganyika), 8.5.7. Ko-na-ne (Hawaii), 4.11.4.
Imperial (Italy, France, England), 6.2.20. Kono (Korea), 4.8.2.
Inaani (NAI.), 6.7.21. Kontjil (Korea), 3.5.23.
Ingilith (Uganda), 8.5.4. Koro (Dogons), 7.5.23.
Intotoi (Kenya), 7.6.20. Korubéddo (Somaliland), 4.4.4.
Intuvigai zuli gairagai (NAI.), 6.7.35. Koti keliya (Ceylon), 5.4.4.
Irish (England), 6.2.11. Koti sellama (Ceylon), 5.5.4.
Isafuba (Mashonaland), 8.5.11. Kotiyo saha harak (Ceylon), 5.6.8.
Ise-ozin-egbe (Nigeria), 7.5.6. Kotra (Iceland), 6.2.39.
Iyagbe (Nigeria), 7.5.54. Kotu-baendum (Ceylon), 7.3.3.
Iyogh (Nigeria), 7.5.52. Kotu ellima (Ceylon), 4.2.30.
Kowwu dunki (India, U.P.), 4.2.37, 8.
Jacina (Italy), 6.2. Kpo (Liberia), 7.5.14.
Jacquet (Spain, France), 6.2.37. Kuwanroku (Japan), 6.6.4.
Jacquet de Versailles (France), 6.2.38. Kvatrutafl (Iceland), 4.1.12.
Jara-badakh (Somaliland), 3.1.10. Kyngee bel (West Africa), page 177.
J erin (Nigeria), 7.5.38. Kyoti (Sahara), 3.6.4.
Jeu-de-dame (England), 4.3.1.
Jodorakase (Gold Coast), 7.5.9. Lam pusri (Assam, Sikkim), 4.2.23.
Jodu (Nigeria), 7.5.9. Lam Turki (India), 2.7.3.
Jungka (Sumatra), 7.4.5. Langa-hola (Dutch Guiana), 7.9.3.
Juroku musashi (Japan), §.2.3. Laquet (Spain), 6.2.21.
Latrunculi (Rome), 2.8.4; 4.1.1.
K‘ab al-baidar (Arabs, Persia), 3.5.13-14. Lau kata kati (Lower Bengal), 4.2.37.
Kale (Gabon), 7.5.21, 29. Leap-frog (England), 4.9.2.
Kaloleh (Sumatra), 7.4.7. Legwet kya (Burma), 5.6.16.
Kanj-guti (Orissa), 7.2.6. Lela (Congo), 8.1.5.
Kaooa (India, C.P.), 5.6.23. Leopard games, $.5.1-5.
Kara (Sudan), 7.5.66. Leyla-gobale (Somaliland), 7.6.6.
Kare, Karnun (Nigeria), 3.6.2. Linpole (France), 6.2.25.
Katra (Madagascar), 8.3.1. Lok (Nigeria), 7.5.39.
Kawada keliya (Ceylon), 6.5.2. Longbeu-a-cha (Assam), 7.2.5.
Kawakib (Persia), 6.8.1. Lontu-holo (Dutch Guiana), 7.9.5.
Kawasukuts (NAI.), 6.7.13. Lorkabéd (Somaliland), 4.4.4.
Kawwa dand (India, U.P.), 2.7.3. Lorzen (Germany), 6.2.25.
Ketchdolt (England), 6.2.4. Lourche (France), 6.2.25.
Keya (West Africa), page 177. Ludo (England), 6.4.16.
Kezute (NAI.), 6.7.35. Luk tsut k’i (China), 3.3.13.
Khutka boia (Punjab), 7.2.1. Lupo e pecore (Italy), 5.3.1, 6.
Kiarabu (Zanzibar), 8.2.2. Lurch (England), 6.2.25.
Kinsgoot (NAI.), 6.7.36. Lurtsch(en) (Germany), 6.2.25.
Kints kut (NAI.), 6.7.36.
Kioz (Arabs, Palestine), 4.10.4. Machanan (Java), 5.6.6.
Kisolo (Congo), 7.8.6. Madell, Sixpenny (England, Wilts.), 3.4.5.
Kiswahili (Zanzibar), 8.3.11. Madji (Dahomey), 7.5.11.
Kit cat cannio (England, Suffolk), 3.2.1. Majnini, Li‘b (Damascus), 7.1.5.
Kiu-kung (China), 3.3.13. Mak khom (Siam), 7.4.1.
Kiusu (Kenya), 7.6.12. Makvar (Denmark), 4.3.5.
Kiuthi (Kenya), 7.6.10-13. Makwinni (Nigeria), 7.5.51.
Ko-app-paw-na (NAI.), 5.1.5. Mak yek (Siam), 4.1.4.
Ko-ko-nag’n (NAI.), 4.11.4. Mak yep (Siam), 4.2.42.
Kolowis awithlaknannai (NAI.), 4.2.31. Malingan (Java), 4.11.2.
 INDEX OF GAMES 253
Malmosdy (Hungary), 3.5.11. Molenspel (Golland), 3.5.6.
Malomjatek (Hungary), 3.5.11. Molino (Italy), page 39.
Mamatyang-matyang (Celebes), 5.6.7. Molle (Denmark), 3.5.7.
Mandoli (Hydra), 7.1.2. Moo (Hawaii), 4.3.4.
Mangala (Arabs), 7.1.9, 10; 7.5.63-64, 67. Morahha (Hanzoan), 8.5.1.
Mangal-pata (Bengal, Behar), 4.2.22. Morris (England), 3.3.6; 3.4.1-4; 3.5.4.
Ma-nu (Hawaii), 5.3.3. Moruba (Bantu), 8.2.17.
Mar (India, U.P.), 4.2.22. Moulin, Jeu de (France), 3.5.2.
Marella (Italy), 4.2.3. Moungala (Nubia), 7.5.67.
Marelle (Italy), 4.2.3; quadruple, 3.7.4; M’raantsou (Comoro J), 8.5.9.
quintuple, 5.3.1. Msuwa (Nyasaland), 8.2.13.
Marelles (England), 5.3.1. Msuwa wa kunja (Nyasaland), 8.2.14.
Marelles (France), 3.3.5. Mu (Hawaii), 4.3.4.
Marene (France, dial.), 3.5.2. Mudrunggum (Deccan), 5.6.18.
Marlin (England, dial.), page 38 n. Mughal-Pathan (Bengal), 4.2.21.
Marquere (Sweden), 4.3.5. Miihle, miihle(n)spiel (Germany), 3.1.3;
Marrella (Sicily), 4.3.1. Br 555:
Marro (Catalonia), 3.3.3; 3.5.13 4.2.2. Miilchenspiel (Germany), 3.5.5.
Marro de punta (Spain), 4.3.1. Mulen, dupelmulen (Germany), 3.5.5.
Matoe (Soemba), 7.4.10. Mulinello (Italy), 3.3.4; 3.5.33 3-7-4-
Mawkar katiya (Assam), 7.2.3. Mulino (Italy), 3.3.4.
Mbangbi (N. Cameroons), 7.5.61. Murrells (England, dial.), page 38 n.
Mbau (Angola), 7.8.8. Musinaykahwhanmetowaywin (NAI.),
Mbéré (A-Ndakko), 7.7.7. Se
Mchombwa (Nyasaland), 8.2.13 n. Mutingwano (A-Kamba), 8.5.3.
Mechiwa (Bali), 7.4.9. Mu-torere (Maoris), 4.8.3.
Mefuhva (Transvaal), 8.2.11. Mweiso (Tanganyika), 8.3.5.
Melnichny (Russia), 3.5.10. Mweso (Uganda), 7.6.17; 8.3.7-9.
Melnitsa (Russia), 3.5.10. Myles (England), 6.2.22.
Mémbille (Mangbattu), 7.7.4. Mylna (Iceland), 3.5.9.
Merafib, El (Sudan), 6.5.8.
Méreles a six tables (France), 3.4.2. Na(h) (Dogons), 7.5.43:
Merelier, Jeu de (France), 3.5.2. Nambidi (Rhodesia), 8.2.9.
Meérelles (France), 3.3.5; 4.2.4. Namudilakunze (Rhodesia), 8.2.18.
Merelli (ML.), 3.5.3. Nao-guti (Punjab), 3.5.17.
Merels, Larger, 2.1.10; 3.5.1-32; Smaller, Naranj (Maldive I.), 7.2.10.
2.110572 .8-15 3-31 —20- Nard (Persia, Arabs), 4.1.3; 6.1.2.
Merls (England, dial.), page 38 n. Nardi (Georgia), 6.1.2.
Merrels, merrils (England), page 38 n. Nardshir (Persia), 6.1.2.
Merry hole, peg (England dial.), page 38 n. Narr (Siberia), 6.1.2.
Merryal (England, dial.), page 38 n. Natatu (Rhodesia), 8.2.7.
Meuchoh, Meuliéh (Sumatra), 7.4.5. Natt klab ash-shawk (Arabs of Palestinc),
Meurimiieng-rimiieng (Sumatra), 5.6.5. 4.2.12.
Meurimiieng-rimiieng-do’ (Sumatra), $.5.5. Nchombwa (Nyasaland), 8.2.5.
Meusuéb (Sumatra), 7.4.5, 6. Nchuwa (Nyasaland), 8.2.6.
Meuta’ (Sumatra), 7.4.5. Ndim-ndum (Nigeria), 7.5.49.
Mig-mang (Tibet), 4.11.3. Ne‘baku‘thana (NAI.), 6.7.31.
Mingma (Siberia), 4.3.12. Negensticken (Holland), 3.5.6.
Minoret(to) (Italy), 6.2.19. Nei-pat-kono (Korea), 4.8.2.
Miracles, miraele (England, dial.), page Nerenchi (Ceylon), 3.1.7; 3.3.12; 3.5.21
38. Nicque noque (France), 6.2.
Mlyn (Bohemia), 3.5.11. Nine holes (Anc. Egypt, England), 2.1.10;
Mogol Putt’han (Deccan), 4.2.25. Bolisihy @
254 INDEX OF GAMES
Niranchi (Ceylon), 3.5.21. Pereaiini (Kenya, Uganda), 8.3.2.
Njombwa (Nyasaland), 8.2.3. Petteia (Greece), 2.7.1; 4.1.1.
Nouo (Nyasaland), 8.2.13 n. Philosopher’s game (Europe), 4.5.2.
Noughts and crosses (Britain), 3.2.1. Picaria (NAI.), 3.3.20; 3.5.27.
Nsoli (Nyasaland), 8.2.5. Pichas (Persia), 6.4.11.
Ntama (Greece), 4.3.1. Piprjall (Iceland), 6.2.40.
Nyout, nyout-nol-ki (Korea), 6.5.4. Pitarilla (NAI.), 3.3.20; 3.5.28.
Nyuwunu bo(uh) ana ya(i) (Dogons), Plaisant (France), 4.3.1.
ESS: Plein (France), 6.2.33.
Poleis (Greece), 2.7.1.
Obridgie (Nigeria), 7.5.36. Po-lo-sai-hi (China), 2.11.
Oca, Giuoco dell’ (Italy), 6.5.5; Juega de Pon chochotl (NAI.), 4.2.18; 5.1.2.
la (Spain), 6.5.5. Pong hau k‘i (China), 4.8.2.
Ochi (Tanganyika), 8.5.6. Poo (Liberia), 7.5.15.
Ofanfelling (Iceland), 6.2.7. Pretwa (India, Behar), 4.2.40.
Officers and men (England), $.3.7. Provincial (England), 6.2.23.
Ogori (Nigeria), 7.5.194. Puff, Lange (Germany), 6.2.44.
Ohohla (NAL), 2.7.3. Puhulmutu (Ceylon), 7.3.1.
Oie, Jeu de l’ (France), 6.5.5. Pulijudam (India), $.5.3.
Okusora (Uganda), 7.6.17. Pu-ni-pe-ki (Hawaii), §.3.3.
Okwe (Nigeria), 7.5.32, 45, 48, 55-8, 65.
Olinda keliya (Ceylon), 7.3. Qamish uyuni (Turkey), 4.10.1.
O-pat-kono (Korea), 4.8.2. Qancebie (Abyssinia), 7.6.4.
Ot-jin (Borneo), 7.4.13. Qirg (Arabs), 3.3.8; page 37.
Oughts and crosses (Britain), 3.2.1. Quecho (Chile), 6.7.3.
Ou-moul-kono (Korea), 4.8.2. Queen’s game (England), 6.2.5.
Ouri. See Wari. Quince (NAI.), 6.7.25, 35.
Owasokotz (NAI.), 6.7.26. Qvarn (Sweden), 3.5.8.

Pachgarhwa (India, U.P.), 7.2.2. Rafaya (India), $.5.4.

Pachih (Sumatra), 6.3.5; 6.4.9. Rafspel (Sweden), 5.3.1.
Pachis (Persia, Ceylon), 6.4.7, II. Real (Spain), 3.5.1.
Pachisi (India), 6.4.6. Rebellionis, Ludus (Italy), 5.4.1.
Padagang (Celebes), 6.1.6. Refskak (Iceland), §.3.1.
Pahada keliya (Ceylon), 6.4.2. Renard (France), §.4.2.
Pa-hi-u-hi-u (Hawaii), 4.11.4. Renard et les poules (France), 5.3.1.
Paitariya (NAL), 3.5.27. Renette (France), 6.2.5.
Pallankuli (Tamils, Ceylon), 7.2.9; 7.3.5. Reverquier (France), 6.2.39.
Pancha (Ceylon), 6.5.1. Reversi (Europe), 4.7.2.
Parchis (Spain), 6.4.11. Revertier (France), 6.2.39.
Pareia de entrada (Spain), 6.2.42. Riga (Italy), 3.3.4; 3.5.3; 4.2.3.
Pasit (Burma), 6.4.8. Rimau, Main (Malaya), 5.6.19.
Pataballang (Celebes), 6.1.5. Rimau kambing, Main (Malaya), 5.6.19.
Patok (Korea), 4.7.1. Rimau-rimau (Malaya), 5.6.1, 4.
Patol (li) (NAI.), 6.7.1, 23, 24. Rimoe (Indonesia), 5.6.2.
Paumecary (England), 6.2.3, 43. Rithmomachy (Europe), 4.5.2.
Pedreria (Spain), 3.3.20. Romanos rencontrat (Spain), 6.2.46.
Pegity (Britain, U.S.A), 3.7.3. Romavoa (NAI.), 6.7.35.
Peki (Congo), 7.8.5. Roseya (Damascus), 7.1.6.
Pendus, Les (France), 3.1.4. Ruweisho (Tanganyika), 8.3.5.
Pentalpha (Crete), 2.7.3; 2.1.10. Ryouk-pat-kono (Korea), 4.8.2.
Pente grammai (Greece), 2.7.3.
Perali kotuma (Ceylon), 4.2.29. Saddigiya (Somaliland), 7.6.5.
 INDEX OF GAMES 255
Sadurangam (S. India, Ceylon), 6.3.4. Solitaire (Europe), 4.9.1.
Sai (Flores), 7.4.11. Solitario (Peru), 5.2.1.
Sake, Len (Siam), 6.1.7. Songo (Cameroons), 7.5.31.
Salta (Europe), 3.8.2. Spreta (Nyasaland), 8.2.10.
Sam k‘i (China), 3.5.22. Srid-pai khorio (Tibet), 6.6.8.
San-noku-narabe (Japan), 3.1.8. Ssang-ryouk (Korea), 6.1.3.
Santaraj (Abyssinia), 3.3.16. Sua tok tong (Siam), 4.8.2.
Sat-gol (India, C.P.), 7.2.11. Sudder, Sudra (Arabs), 3.5.13.
Satoel (Indonesia), 4.2.36. Sugoroku (Japan), 6.1.3.
Saturankam (S. India, Ceylon), 6.3.2. Sujjua (India, U-.P.), 3.5.18.
Sbaraglino, Sbaraglio (Italy), 6.2.13, 15. Sullusie (Abyssinia), 7.6.3.
Sbaraia, Sbaraino, Sbaraiono (Italy), Sunoroku (Japan), 6.1.3.
6.2.14. Sute (Nyasaland), 8.2.19.
Schaap en wolf (Holland), 5.3.1. Syredge (Nubia), 4.1.11.
Schafzagel (Germany, dial.), 3.5.5; 4.1.13.
Schuster (Sweden), 6.2.36. Tab (Egypt, Arabs), 4.10.1.
Senat, sen’t (Anc. Egypt), 2.1.1, 6; (Ur), Tab al-qasab (Arabs), 4.10.1.
DPS Tab wa dukk (Arabs), 4.10.1.
Serup (NAI.), 6.7.34. Tabal, Main (Malaya), 6.1.4.
Set-dilth’ (NAI.), 6.7.16. Tablas (Spain), 6.2.12.
Set-tilth’ (NAI.), 6.7.18. Tablas, Las quinze (Spain), 6.2.10.
Seys, dos e as (Spain), 6.2.24. Tablé (Greece), 2.8.3; 6.1.1.
Sgjoug (Armenia), 3.5.16. Tables (France, England), 6.2.
Shagchakar (Assam, Sikkim), 5.6.22. Tables avalées, rabbatuées (France), 6.2.6.
Shah (Somaliland), 3.5.25. Tablut (Lapland), 4.1.13.
Shantarad (Somaliland), 4.1.8. Taboas (Portugal), 6.2.
Shanzdahu-kattaru (Turkestan), 4.2.20. Tabula (Rome), 2.8.3; 6.1.1.
Shap luk kon tseung (China), 5.2.2. Tefl (England), 4.1.13.
Shashki (Russia), 4.3.1, 6. Tafl (Scandinavia, Iceland), 4.1.13.
Shatranj shir bakri (India), 5.4.3. Tami (Gold Coast), page 177.
Shayo (Nigeria), 7.5.17. Tammam (Somaliland), 6.3.6.
Shepherd’s mill (England), page 39. Tant fant (Lower Bengal), 3.3.10.
Sher bakar (Punjab), 5.6.11. Tapal empat, Main (Malaya), 5.6.3.
Shi liu kan tsiang kiin (China), 5.2.2. Tapatan (Philippines), 3.3.14.
Shimunana (Rhodesia), 8.2.12. Tasholiwe (NAL), 6.7.20.
Shing kun t’o (China), 6.6.1, 2. Tatsungin (NAI.), 6.7.5.
Shiva (Nigeria), 3.6.2. Tau (Anc. Egypt); 2:1.3, 7, 8:
Sholiwe (NAI.), 6.7.29. Taule (Provence, France), 6.2.
Shwan-liu (China), 2.11; 6.1.3. Tavel (England), 6.2.
Sidere (Persia), 3.5.14. Tavola (Italy), 3.5.3.
Siga (Egypt), 4.1.7; 4.4.3; (S. India, Cey- Tavola da molino (Italy), 3.5.3.
lon), 6.3.2, 4; (Senegal), 4.10.5. Tavole (Italy), 6.2.
Sija (Sudan), 4.1.9. Tavole reale (Italy), 6.2.12.
Sijat el taba (Sudan), 4.10.3. Tavoletta (Italy), 3.3.4.
Sipchi kat (Assam, Sikkim), 4.2.23. Tawfa (NAL), 6.7.224.
Si-perde (Persia), 3.5.14. Tawlbwrdd (Wales), 4.1.13.
Sitih (NAI.), 6.7.17. Tawula (Turkey, Egypt), 6.1.8.
Sitta (Hanzoan), 4.10.1. Tayam sonalu (Tamils), 6.4.4.
Smerelli (Italy), 3.3.4; 3.4.1; 3-5.3- Terhiichii (Assam), 4.2.11-31.
Snake (England), 6.5.6. Testa (Italy), 6.2.25.
Snakes and ladders (England), 6.5.7. Thshan-han (China), 2.11.
Sokkattan (Tamils), 6.4.7. Ti (Sierra Leone), 7.5.26.
Solah guttiya (Bengal), 4.2.21. Tichha (India, U.P.), 4.2.22.
256 INDEX OF GAMES
Ticktack (England), 6.2.29. Um el-banat (Sudan), 7.6.8.
Tick tack toe (England, Warwick, Staffs.), Um el-tuweisat (Sudan), 7.5.64.
B23. Uno-alti-tashu (Turkestan), 4.2.20.
Tieste (France), 6.2.25. Uruheisho, uruheso (Tanganyika), 8.3.4.
Tiger games, 5.6.1-23. Uthi (Kenya), 7.6.16.
Tik tak tol (Holland), 3.2.2. Uugg (Abyssinia), 7.6.2.
Til guti (India), 7.2.7.
Tin-guti pait pait (Bengal), 3.3.11. Vai lung thlan (Assam), 7.2.4.
Tip tap toe (England, Lincoln), 3.2.1. Vargen och faren (Sweden), 5.4.4.
Tit tat toe (England, U.S.A.), 3.2.1. Vartafvel (Sweden), 6.2.
Tjyong-kyeng-to (Korea), 6.6.3, 6. Vela (Madagascar), 4.6.1.
Tobi-shnogi (Japan), 4.4.2. Verkeer (Holland), 6.2.39.
Tobi-sugoroku (Japan), 6.6.5. Verkehren (Germany), 6.2.39.
Toccadiglio, toccateglio (Italy, Germany), Verkehren utan junker (Sweden), 6.2.41.
6:22:35; Verquere (England), 6.2.39.
Todas tablas (Spain), 6.2.11. Volki ovtsy (Russia), 5.3.1.
Tods and lambs (England, dial.), 5.3.1. Vrhcaby (Czech), 6.2.
Togatog (Somaliland), 7.6.5.
Tohi (Bongo tribe, Cent. Africa), 7.7.2. Wakanpamungthae (NAI.), 4.3.2.
Totolospi (NAI.), 4.1.15; 6.7.9, Io. Walak-pussa (Ceylon), 7.3.2.
Tourne-case (France), 6.2.9. Wali (Sahara), 3.6.4.
Toutes tables (France), 6.2.11. Walu (Dogons), 7.5.22.
Tre-guti (Punjab), 3.1.6. Warcaby (Poland), 6.2.
Tria (Italy), 3.5.3. Wari (W. Africa), 3.6.1; 7.5.12, 27, 35, 57:
Trictrac (France, England), 6.2.29. Wari, English (Barbados), 7.9.2.
Trictrac, Grand (France), 6.2.30, 31. Wari, French (Barbados), 7.9.1.
Trictrac 4 ecrire (France), 6.2.32. Wartabl (Denmark), 6.2.
Triodi (Macedonia, S. Greece), 3.5.12. Wash‘kasi (NAI.), 6.7.33.
Triodion (Greece), 3.5.12. Wasokuts (NAI.), 6.7.26.
Trip(p) trap(p) trul(1) (Saxony, Sweden), Wei-k‘i (China), 4.7.1.
312.3% Werptafel (MDu.), 6.2.
Trique (Amazon), 3.5.32. Wet (Shilluk), 7.7.3.
Tshuba (Rhodesia), 8.2.12, 17. Whyo (Nigeria), 7.5.37.
T‘shu-p‘u (China), 2.11. Wolo (Ivory Coast), 7.5.15.
Tsi(n)dil (NAL), 6.7.19. Wora, woro (W. Africa), 7.5.13, 15.
Tsikehi (NAI.), 6.7.8. Worptafel (MDu.), 6.2.
Tsona (NAI.), 6.7.31. Wurfzabel (Germany), 4.1.13.
Tsoro (Mashonaland), 8.2.15. Wu-sho (China), 2.11.
Tsun k‘i (China), 6.5.9.
Tubn, tubna (Arabs), 3.5.13. Yakamaido (NAI.), 3.5.31.
Tugea (NAI.), 6.7.23. Yasasukari musashi (Japan), 5.1.6.
Tugi-e-pfe (NAI.), 6.7.25. Yeung luk sz‘ kon tseung kwan (China),
Tuknanavuhpi (NAL), 4.2.19. 5.2.4.

Ufuba wa hulana (Hanzoan), 4.1.10. Zamma (Sahara), 4.2.14, 33.

Um el-bagara (Sudan), 7.5.63. Zohn Ahl (NAI.), 6.7.30.
 INDEX OF GAMES 257

GAMES OF WHICH THE NAMES ARE UNKNOWN

Abyssinia, 4.2.13. Liberia, 4.4.7.
Annam, 7.4.2. Madagascar, 4.2.15.
Assam, 3.5.19; 5.6.12, 17, 343 4.2.9, 10, 24. Mongolia, 5.6.14.
Cameroons, 7.5.30. INDANGs SEGHOR Sines CH, Cy 9 1, PT Bp
Ceylon, 6.5.3. 37-
GhinayS;5)7-4.3- Nigeria, 7.5.53, 60.
Congo Basin, 7.8.3, 4; 8.5.8. Somaliland, 8.1.3.
England, 4.2.6. Switzerland, 4.2.5.
India; Madras, 7.2.8; Orissa, 5.6.20 Tanganyika, 8.3.6.
South, 3.5.20; U.P., 4.2.8. Tibet, 6.6.7.
Indonesia, 4.2.28. Uganda, 7.6.9, 18, 19; 8.5.5

5356
 INDEX OF "GAMES ARKRANG EDSDy
COUNTRIES
AFRICA
Abyssinia, 3.3.16; 4.2.13; 7.6.1-4; 8.1.2. 3.6:2, AsnO:4.LOss 725-05 7p 95 L720, 0325
Angola, 7.8.7, 8. 34, 36-42, 45-61.
Ashanti, 3.6.1. Nubia, 4.1.11; 7.5.67.
Bahr el-Ghazal, 7.7.1-8. Sahara, 3.6.4, 5; 4.2.14, 353 4.4.6.
Cameroons, 7.5.30, 31, 62. Senegal, 4.10.5; 7.5.1.
Comoro Islands, 8.5.9. Sierra) Leone, 7.5.20; 27+
Congo Basin: 7.8.1-6; 8.1.5; 8.5.8. Somaliland; 3.1-10; 3:3.173 3.5.25; 4-1-8
Dahomey, 4.10.6; 7.5.10-12, 16, 33, 35. 4.4.43 4.10.2; 6.3.6; 6.4.13; 7.6.5-7.
Dogons, French West Africa, 7.5.3, 22-25, South Africa, Achikunda, 8.2.10, 19;
43, 44. Angoni tribe, 8.2.5; Atonga, 8.2.6;
lea, Shite OR big, 28 vinteyiis Crsiiste Ba-lIla tribe, 8.2.7-9, 18; Bantu, 8.2.17;
TAA, 7, LO BaThonga, 8.2.12,17; Ba Venda, 8.2.11;
Gabon, 7.5.21, 29. Damara, Herero, 8.2.4; Manyanja,
Gambia, 4.4.5; 7.5.13, I5. 8.2.13, 14; 8.4.1, 2; Mashona, 8.2.15;
Gold Goast,, 371.1P33 -4:4s) 355.205a se Os tes 8.5.11; Yao tribe, 8.2.3; 8.3.11.
7s5.Oy 025130: Sudan,.3:6-33.4-1.93/4-10:31510:5:85 7-522005s
Hanzoan, 4.1.10}; 4.10.1; 8.5.1. 8, 63, 64, 66; 7.6.8; 8.5.2.
Ivory Coast, 7.5.15. Tanganyika, 8.3.4-6; 8.5.6, 7.
Kenya, 7.6.10-16, 19, 20; 8.3.2. Togoland, 7.5.28, 36.
Liberia, 4.4.73; 7.5.14, 15. Uganda, 7.6.9, 17, 18; 8.1.4; 8.3.3, 7-10;
Madagascar, 3.3.19; 4.2.15; 4.6.1; 8.3.1; 8°5:4,,5>
8.5.10. Urindi, 8.3.4.
INigeriay 69.0.0 0393-320 0573-404 3325.20 Zanzibar oral 2o 3.10s

NORTH AMERICA
Canada, 4.3.8. 6.7.1-36.
NAL, 2.7.3; 3.3-203 3.5.27-313; 4.1.14-153 U.S.A., 3.2.1; 6.2.48; 8.2.16.
A 21O—-19,5 A3%, 4.3.25, 5.1-2—-$5 523225 West Indies, 7.9.1, 2.

SOUTH AMERICA
Amazon, 3.5.32. Dutch Guiana, 7.9.3-S.
Chile (Aracaunians), 4.11.1; 6.7.3. Rerumes-2en-

ASIA
Annam, 7.4.2. Indias General. 525295.4510-3-00-401513 51400:
JPL SES BRIS Boas US IVI, rene IASsain 33521920422. O-1102 35 eae
6.8.1; 7.1.1, 9, 10, II-14; 8.1.1. 5.0.0 259075 225 2350 722355
Armenia, 3.5.16. Behar, 4.2.40.
Burma, §.6.16; 6.4.8. Bengaly2.72313-4.LOpulers sel neatols
@evlonys3.8-9)563-3-
02 sass 2 4200s 225 372 5.0.921053-3-
303 4.3.93 5.4.45 5-5-4; 5.6.3, 8; 6.3.4; Central Provinces, 4.2.22, 38, 39, 41;
6.4.2, 7, II, 15; 6.5-1-3
3 7-3-1-S. S0123/20 7-2-01.
Olvisey Beep Caos Mens COR Spy), Chota-Nagpur, 7.2.7.
A O2:1.3'> 0: 5.0,0:0:0 0207-4 3 Weccany4 225 7.5:6:15 e118.
 GAMES ARRANGED BY COUNTRIES 259
India (cont.). Iraqy Galelerse
Madras, 6.4.14; 7.2.8. eli, Syn7id Zee 7S foe CM ROR (ines
Orissa, 5.6.21; 7.2.6. 6.8.1.
INbeploy gigi? Spi (lore, Bp Pete apatiyy 3-0-830 3-720s 04kO,md 4.2m Ls
SOS Dele S21:028922-3)3) On-3'7 023-531 0O-4sn Se
Sa lndiah3s5220310:35259457410.5.0: Koreas 3252235 407.Lard Solin 23) Ones) 0-54
Tamils, 6.4.4, 7; 7.2.9. 6.6.3, 6.
United Provinces, 2.7.3; 3.5.18; 4.2.8, Moalavas Antss 422.27 0Ae3 0s 4-0195552 000s
2293 OES. OnlOnn 722.24 By hy OG eS LEO Gyles
Indonesia, Achehn, 4.2.33; 5.5.53 5-6-5} Maldive Islands, 6.4.10; 7.2.10.
OFA.) 7 ALS; O: Palestine (Arabs), 3.5.13; 4.2.12; 4.4.1;
Bali, 7.4.9. 4.10.43 6.4.12.
Borneo, 7.4.12, 13. Philippines, 3.3.14; 7.4.16, 17.
Celebes, 4.1.12; 5.6.7; 6.1.5,
6; 7.4.14, 15. Siamiyedeied sedi2 Aa Are. oases Oskar
Flores, 7.4.11. 6.4.18; 7.4.1.
Java, 4.2.34; 4.11.2; 5.6.6; 7.4.8. Siberia, 4.3.12; 5.6.13.
Saleijer, 4.2.28. Sy1ayOF4.0'7.217. 1255 Osic>
Silamur, 4.2.36; $.6.2. Mubet, 4s0l.310:0:75 0
Soemba, 7.4.10. Turkestan, 4.2.20.
Sumatra, 7.4.5, 7- Muarkeys 3:5.155)4.3615 4-4-4054. 10. + On0-0-

EUROPE
Generally, 3.5.1-12; 3.7.2-4; 3.8.1, 2; Hungary, 3.5.11.
Ae] :2i30A-OeL 5 523-79055 Iceland) 355-0374.
1- 1335-328 359367 O12 aes
Czechoslovakia, 3.5.11. 39, 40, $2.
Denmark, 9355075) -463-05 53563071 012-393 Ireland’ 2:0:1, 23 3:3:73)4e1.2-
OOS.S: Italy) 333545 3-4-5599) 44-503 307s 42s
Enolandyssst-i) 92250; 343. 0514-4055) 325245 Ac Qulls 33 4eSe2g Sac lgOs5 740 3 Os2eL i054;
AnlelgntA Qs Aro ren LO; Dies Aase2y 4.0.2 26, 35, 37, 425 45, 49, 515 6.5.5.
5-3-1, 4, 5, 7; 6.2.3-5, II, 12, 15, 18, 20, Lapland, 4.1.13.
23, 25, 29, 39, 43, 44, 475 6.4.16; 6.5.6, 7. Portugal, 4.3.1, 4.
Finland, 4.3.1. Russia; 3552105) 4:3.1,1035.3.0-
ETANCensseleA mgs
eSes ede 5 BsSs2 ss e7eds Scotland, 3.3.6; 4.3.1; 6.2.12.
M2-AS A431, Fp 104. 5:27°5. 3.0, 73 Sed 2s Spain, 3.3.2, 3, 20; 3.5.1; 4.2.1, 23 4.3.45
6.2.5, 6, 9, II, 18, 25, 26, 28-32, 37-39; S-Ueds Os2aly 2p LO—12,, LO—18,) 21,2425,
6.5.5. 27, 37) 42, 45, 46; 6.8.1.
Germany-sal.ssio 5-5 4G4asels |seals Swedense3.2-95 355-83) 4-sel eSeesiaelmys
73 6.2.25, 35 39, 44, 47, 50; 6.5.5. 5.4.45 6.2.36, 39, 41.
Greece m2 7 ae srs oni A sso my ekeos Ukraine, 4.3.1, 6.
Holland yestte2snoee
2 as 5 On Aor soe tens Wales, 2.9.1; 4.1.2.
6.2.39; 6.5.5.

PACIFIC ISLANDS
Hawaii, 4.11.4; 5.3.3. | New Zealand (Maoris), 4.8.3.
 GENERALSINDEX
Abbott, G. F., 47. Beale, F., 102.
Abbott, W. L., 201. Becq de Fouquiéres, L., 32.
Abdallah Shah, 206. Beham, H. S., 44.
Abu Habbah (Sippar), 22. Belgrave, C. D., 83.
‘Adli, al-, 115. Beni-hassan, 13.
fEthelstan, 61. Bennett, C. R., 103.
Agathias, 32. Bennett Ge len asie
Agra, 132. Bent, J. T., 194, 195, 205, 206, 225.
Agricola, J., 119. Bentley, H., 201.
Ak-hor, 17, 18, 34. Bérenger-Féraud, 83, 225.
Alcaeus, 28 n. Bergh, V. d., 201.
ALBA, 29, 31, II3. Best, Elsdon, 93.
Alfonso X, 2. Beurt, Ch., 174.
Alfonso MS., 40, 41, 82, 95, 98, 99, 118, Bharhut, 130.
NOLO), TT MP, TPE MI, TEL, TRIG, Birket-Smith, E., 232 n.
Allahabad, 132. Bligh, Capt., W., 238.
ALQUERQUE: boards, 6; name, 36. BLOCKADE-GAMES, 92.
Alvarez, Fr., 194, 195. BLOT, 120.
AMERICAN INDIAN GAMES, NATIVE: war, Boarp-GaMes: defined, 1; classified, 4;
OAR 71s TACOMATec230. diffusion, 229, 230; distribution, 229,
Ancient Laws of Wales, 63. 230.
Andrée, R., 185. Boarops: latticed, 6; lined, §; incised in
Andrews, 97. castles, 44, 102; in choir-stalls, 41, 44; in
Apollo Shroving, 43. cloisters, 39, 41, 44; on tomb-stones, 41,
Approach captures, 11, 88. 44, 66; built into walls of churches, 41,
Araqi, 115. 44; combined, 59 n.
Arber, E., 142. Boghaz-K6i, 22.
Arbory, 41. Bolton, H. C., $4.
Ardshir (Artaxerxes), 114. Boniface, Pope, 35.
Aristotle, 27. Bonus Socius MS., 3, 45, 118, 120, 123, 124.
ASHTAPADA, 129. Book of Lismore, 34.
Athens, 26. Books of the Dead (Egypt), 14.
Austin, RG. 24, 285209 n\, 30n., 32, 33. Bouche, O., 201.
Avelot, R., 83, 160, 177, 178, 179, 181, Bowdichy le best 7 ie
182, 184, 185, 194, 199, 200, 201, 206, Bown o Hamtwn, 62.
224. Box-boards, 16, 20, 44.
Ayrton, E. R., 12 n. Boyden, E. C., 122, 129.
Boy’s own Book, 46 n., 104, 117.
BACKGAMMON: origin of name, 122; mean- Brahma-jala sutra, 35.
ing of after-game and back-game, 122. Brand, J., 39.
Ballinderry, 59. Braunholtz, H. J., 184, 219.
Barclay, A., 120. British Chess Magazine, 45 n., $0.
Barozzi, Fr., 84. Brockhaus, 79.
Bartinelli, M., 123. Brown, Mrs. W. W_ 78.
Bartsch, B. P., 119. Browne, E. G., 113.
Basden, G. T., 197. Brugsch, H. K., 13 n
Basinstoke, 58. Brunets i... 185.
BATTLE-GAMES, 53. Brunet y Bellet, J., 41, 65, 105.
262 GENERAL INDEX
Bullock, C., 212. Corvinus, G. S., 44.
Burton, R., 84. Cotgrave, R., 42, 76, 82, 121.
Burton, R. F., 136, 195. Cotton, C., 120, 121, 122, 124.
Biittikofer, J., 83. Country Life, 39.
Buzurgmihr, 114. Covarrubias, 65.
Coxn@aptai 234.
Caillie, 96 Cr Bri Chualann, 44.
Cambridge Papers, vi, 169, 184, 195, 217. Craftsman, 80.
Cameron, V. L., 201. Gretes2373128:
Campbell, D. M., 68, 91, I10, 175. Crofton, T., 41.
Canalejas, J. G., 75, 78. Cubbon, W., 39 n.
Canterbury, 41. Culin, Stewart, 3, 29, 42, 47, 48, 64, 67,
Captures: classified, 10. 71, 90N., 92, 94, 97, 99, 100, IOI, 104,
Cardan, H., 3, 119, 122, 126, 129. 107, II1§ n., 116, 131, 134, 135, 136, 138,
Cardinall, A. W., 49 n., 176, 185. I4I, 142, 144, 146, 147, 148, 149, 150,
Carnarvon, Lord, 15 n., 16. ISI, 1§2, 153, 154, 155, 156, 158, 160,
Carrerane Aly dane AO OS. 0A2" 167, 168.) 170. 174,017 0,7jentole tOz:
Carsten, 39. 183, 185, 201, 212, 224, 225, 232, 233,
Carter, H., 16 n. 234.
@asanas, bavi bs aS De Cunningham, A., 130.
Cavendish, 66. Cushing, F. H., 64.
CELTIC GAMES, 34. Cyprus, 23.
Champollion, J. F., 13 n.
Chanson de Roland, 119. Dale, A. M., 210, 212.
Chao Wu King, 89. d’Amerval, Eloi, 76.
Charms, 6, 237. Danielli, Mrs. M., 42, 88, 206.
Chatelain, Héli, 201. Dartnell, G. E., 43.
Chatrang-namak, 114. Datta; jal 41,.075770y LLO:
Chaucer, G., 74, 75. Davies, R., 49, 55, 96, 142, 179, 180, 194,
Chedworth, 33. 196.
Cheltenham Examiner, 81. de Boissiére, C., 84.
Cuess: diffusion of, 83, 228. Deir el-Bahri, 16.
CHILDREN’S GAMES, 39, 40, 41, 43, 60, 82, de la Condamine, C., 80.
136, 143, 167, 174, 180, 195, 206, 215, Delafosse, M., 181.
223, 224. d’Escamps, H., 224.
Chitupada Sila, 130. Destruction of Troy, 75 n.
Cho-Yo, 50, 82. Dice: European technical terms, 118.
Cilurnum, 33. DICE-GAMES, 94, 95; 113 ff.
Citolini, A., 118. Dihya, ibn, 74.
Civis Bononiae MS., 3, 42, 45, 46, 118, 120, Divination, use of games for, 50, 235.
123, 124, 128, 129: Dogs (game-pieces), 114.
Claudius, 31. Doughty, C. M., 159, 165, 168.
CLEARANCE GAMES, 93. Dover Castle, 44.
Cohen, M., 160, 205, 206. Dozy, R. F. A., 41.
Colt, Sir G., 66. Drake-Brockman, R. G., 48.
Combined boards, $9 n. Drawn games, 227.
Confucius, 36, 42. Driberg, J. H., 159, 164, 214, 215, 216.
Cook, Capt. J., 97. Dubois, 66, 87.
Corbridge, 33. Ducange, C. D., 42.
Cormac, 34. Du Chaillu, ror.
Corney, 97. Dufresne, J., 80.
Corré, 225. Duibhinnsi, Bishop, 61.
 GENERAL INDEX 263
Duncan, J., 181. GaMgs: antiquity of, 1, 12; athletic, 233;
Durai, Mrs. H. G., 169. canonical, 235; sacred, 235; string, 238;
Duran, D., 147. played at new moon, 235, at the New
Dussek Owl, 109; 137. Year, 170; at wakes for the dead, 235.
GAMES OF CONFIGURATION, $0.
Egharevba, J. U., 180, 190, 191. GAMES ACTUALLY PLAYED: specimen, 45,
Egypt: pre-dynastic, 12; dynasties, I, 14; 89, 108, 109, 182, 186, 188, I9I, 193,
EXT S pLOse Nl, 13 LOS Nav LL eTOn eNee PO lshyiM Altres PIP),
14. GAMMON (Scotch): origin of name, 122.
Eleusis, 26. Garcez | @2.0 78.
Elura, 35. Giethen, L., 185.
Elvira, 32. Gizeh, 161.
Embden, E. van, 81. Gloucester, 41, 102.
Embley, E. B., 224. Goddard, A. R., 26n., 48.
Emery, W. B., 30n. Gokstad ship, 58.
Encyclopédie méthodique, 113, 121, 142. Golberry, 177, 179.
Encyclopedie van Nederlandsch Indie, 174, Gomme, A., 40, IOI.
175. Good, Rev. A. C., 183.
Engelhards, H. E. D., 68. Gorlitz, R. A. V. W., 84.
Esar-haddon, 22. Gotama, 35.
Eskimos, 4, 232. Grace (in race-games), 132, 135, 136.
Eustathius, 26. GRAND TRICTRAC, I25, 126.
Greece, 24 ff.
Faber Stapulensis, J., 84. Greenland Lay of Atli, 60.
Balkenenmyeb.s1 545) 10,13 0.013450 3iSe Grettis saga, 101.
Faraj, Abu’l-, 165. Griaule, M., 42, 66, 180, 183, 187.
Fatehpur-Sikrim, 132. Grimes, W. F., 30 n.
Faversham, §1. Grindley, Miss, 207.
Fejervary Codex, 147, 148. Groos, K., $3, 234.
Felkin, E. W., 180. Groves, Capt. E. T. N., 216.
Field, The, $4. Guest, Lady C., 30.
Finn Magnussen, 63. Gunung, 68.
Firdawsi, $4, I14. Gupta, H. C. D., 39, 47, 66, 67, 68, 70, 71,
Fiske, Willard, 3, 39, 40, 41, 45, 46, 47, 50, O25 Til Tl2, O00} B70:
TOL. «1O$y 150: 123, 020, 1275 128) 120.
FIvE-IN-A-ROW GAMES, $0. Haddon wAn Ga nlo4aause soe
Fletcher, R., 187. Haddon, Kathleen, see Rishbeth, Mrs.
Florio, J., 82. Hahn, A. v., 45.
Foa, E., 181. HALA-TAFL, IOI.
Forbes, D., 97, 147, 234. BAU IR, GIS 74.5 ALE
Fornaldar Ségur, 61. Hamy, Dr., 178.
Fournivall, Richard de, 3. Hanifa, abu, 205.
Fridpjof’s saga, 61. Hargrave, 21, 44.
Fulke, W., 84. Harun ar-Rashid, 115.
Hatshepu, Queen, 16, 18.
Gadd) Gs), 16n., 21,23 1. Hellier, M., 172.
Gaimaediyagala, 160. Helmsley Castle, 44.
Gaimar, G., 62. hlerklotss GaAs 639-1 Tr ruisceT sige
Gallehuus, 60. Herodotus, 8.
Galt, J., 158, 166. Herskovits, M. J., 159, 185, 202, 203, 235.
Game-manuals, 3. Herverar saga, 60.
Game-materials, 2. Hesychius, 26, 27, 28.
264 GENERAL INDEX
Hilmi Samara, 66, 82, 136. K (MS. Brit. Mus. King’s Lib., 13. A.
Himly, K., 36, 42 n., 47, 99, 100, 113, 116. xviii), 118, 120, 123, 124, 128.
History of Chess (H. J. R. Murray), I n., Kadri, M. S., 116.
I3n., 44, 56, 64, 83, 129, 130, 146, 156, Karnak, 18, I61.
234, 235. Kaudern, W., $5, I10, 116, 175, 176.
Hobley, C. W., 197, 206, 224. Kear, J. A., 82.
Hodson, T. C., 47, 66, 67, III. Khusraw Parwiz, I14.
Holme, R., 77, 103, 104, 120, 121, 142, King, Capt. J., 97.
143. Kirkby Underdale, 44.
Holt, 30. Kitab al-aghani, 37.
Holzhausen, Freiherr v., 227. Kitching, A. L., 159.
Horay Sa, 29.0, 67112: Klose, H., 184.
Howell, J., 122. Knossos, 33.
Hoyle, E:, 122. Kollmann, 160.
Huff capture, 11, 65, 69, 81, 105. Komaroy, P., 67.
Huizinga, J., 164, 236, 237. Kraitchik, M., 39, 50, 179, 227 n.
Humphries, E. de M., 29 n., 41, 47, 67, 70, Kroéka-Refs, 58 n.
IIO, 169. Krossmylna, 45.
Humpidge, K. P., 182. Kurna, 18, 39, 42, 44, 54, 65, 160, 161.
Hun Tsun Siti, 36.
Hunfalvy, J., 201. Labat, Ro P.5 177:
Hutton, J. H., 66, 68, 111. La Borde, L. de, 41, 44, 120.
Huxley, Elspeth, 199. Lahun, el-, 16.
Hyde, T., 3, 32, 38, 39, 41, 42, 43, 44, 46, Lala Raja Babu, 106.
ATs S45. 55s1 S25 953.90; LOL, LIZ, 115,008, Lallement, J. G., 79.
I2I, 133, 144, 146, I6I, 165, 166, 206, Lamer, Hi, 29'n-
224. Lanciani, R20:
Lane; E. W., $2, 95, 167-
Ickford, 44. Lap (in mancala games), 163.
Illustrated London News, 15, 18, 220 n. Larkin, Major P. M., 224.
Ingrams, W. H., 207, 208, 220. Laud, J., 77.
Interception capture, 10, 27, 53, 54, 55, 64, Laus Pisonis, 29, 33.
92, 100, IOI, 232. Ravyardy Sine AL.22.
Intervention capture, Il, 54, 100, IOI. Leakey, LS: Bi 160,197, 198:
Intolerance of games, 32, 33 n., II9. Leap captures, 11, 65, 69, 71, 82, 101; line
Isertnes B.a7i7s leap, 108.
Isidore, Bishop, 29, 31 n. LEOPARD GAMES, 106, 107.
Islenzkar Gatur, 47, 103, 121, 129. Lepsius, C. B., 13 n.
Ivan IV, Tsar, 33 n. Lewis, F. R., 55, 63 n.
Liddell, D. M., 26 n.
Jackson, H., 32. Lindblom, 198.
Jacobson, E., 70, 109. Linnaeus, $56, 63.
Jacobson, S., 175. Linton, R., 214.
Jeu de Trictrac, 121, 126. Lister, 47.
Jobson, R., 177. ibopte, NW Ik, Sys seo
Johns, C. N., 44 n. Lots, 5, 7-9; long dice, 8, 237; cubic dice, 8.
Johnston, Sir H. H., 158, 201, 230. Louis IX, 119.
Jorgensen, S. A., 79. Louis XIV, 125.
Joyce, TA 206,224 Lovelace, R., 102.
JuncdHwA, ort, 212. Low, Capt. J., $4, 71, 106, 111, 116, 138,
Justinian, 32. 173.
Juvenal, 33. Ieticas sel 7772
 GENERAL INDEX 265
Lucena, 78. Murray, Miss K. M. E., 66, 92, 102.
Liders, H., 35, 237. Museum fiir Vélkerkunde, Berlin, 131,
Ludlow, 41. 136, 137, 175.
Ludovisi, L., 47, 68, 106, 140. Museums: Aberdeen, 177; British Mu-
Lumbholtz, C., 175. SEIS) 83) 1555 205225) 3955 Os 7 Ase O55
LuRCH: origin of name, 122, 124. 224; Brussels, 12; Cairo, 16, 17; Chesters,
Luschan, F. v., 160, 166, 170. 33; Constantinople, 16, 22; Ipswich,
Lutterhold, Mrs. K., 184. 189; Ireland, 59; Minusink, 111; New
Luxor, I9, 161. York, 16, 21; Oxford (Ashmolean), 6;
Paris (Louvre), 16n.; Paris (Musée de
Mabinogion, 34. ?Homme), 182, 206n., 215; Pennsyl-
Macalister, R. A. S., 22, 23, 44. vania, 21, 64, 67, 113; Richborough, 33;
McGrnigor, Sir J., 177. South Kensington (Victoria and Albert),
Macrobius, 34. 44, 66, 104; Wales, 30; Washington,
Magyar, L., 201. 201, 207.
Mahabhashya, 35. Myres, Sir J., 226.
Mahasna, el-, 12, 235.
Maison des Jeux académiques, 3, 103. Neville, G. Archbishop, 84.
Mallet, P., 75, 76, 78, 81, 106. Newberry, R. J., 182, 187.
Ma’min, al-, 115. Niebuhr, H., 39, 41, 47.
Mandra, 33. Nineveh, 22.
Manoury, 80. Noldeke, Th., 27 n.
Marin, G., 16n., 40, 48, 55, 83, 96, 131, NOMENCLATURE: evidence for diffusion of
140, 141, 195, 196, 234. games, 228; general European prin-
Marmadux, SirJ., 118. ciples, 25; Latin, 29.
Marmion, S., 102. Nopa, 14 n., 40, 43, 48.
Matthes, B. E., 55, 110, 116, 175. Norfolk, Duchess of, 142.
Meare, 8, 237 n. Norwich, Cathedral cloisters, 41, 66;
Meek, C. K., 197, 235 n. Castle, 44, 102.
Megiddo, 23. Niashirwan, $4.
Meldrum, Hon. D., 94.
Mencius, 36. Oluwo, Afin, 192.
Merker, C., 199. Opinion publique, L’, 81.
Michel, C. B., 206. OPTATIVE GAMES, 118.
Migeod, F. W. N., 182. Oraibi, 65.
Mockler, Capt. A. F., 183. Orde Brown, G. St. J. J., 198.
Molina, 97. Orleans, Philippe II, 80.
Monod, T., 49, 66, 69. Ostia, 31.
Monrad, B. O., 177. Overbeck, H., 174.
Montag, J. B., 79, 80. Overbergh, M. B., 200.
Montelius, $7. Ovid, 29, 30.
Montero, P. R., 79. Owen, S. G., 29 n.
Montgomery, Rev. W., 66, 67.
Mosso, A., 24 n. Pallas; PaS:, 82:
Moulidars, 41, 46, IOI, 104, 127. Pallebaedda, 160.
Mousket, P., 74, 75. Panhuys, L. C. van, 202.
Moves: classified, 9. Panton, Colonel, 77.
Murray, A. S., 23 n. Parker, El) 18; 395 41,.47,154, 072108, 1825
Murray, K. C., 14 n., 40, 42, 43, 48, 49, 97, 82, 83, 96, 106, 107, I10, 130, 131, 134,
1335 LOL II7s 1755 LOlselol los kos> 135,, 140, 160) 108; 170,172; 1735) L775
186, 187, 188, 189, I91, 192, 193, 194, 182.
206 n., 235. Parry, N., 169.
266 GENERAL INDEX
Partridge, C., 160, 187, 189. Rodd, F. R., 49.
Passarge, S., 187. Rohlfs, 185.
Payne, W., 77. Rome, 29 ff.; San Paolo, 102.
Pendlebury, J. D. S., 24. Round (in mancala games), 164.
Pensthorpe, 56. Running Jenny, 45.
Pentagram, 19, 28.
RercyaReandiS..555: Saffron Walden, 45.
Perugia MS., 103, 104. Sakkara, 15.
Peter the Great, Tsar, 33 n. Salisbury, 61.
Petrie, Sir W. M. Flinders, 15 n., 16. Salisbury, John of, 102 n.
Petroff, A. D., 79, 81. Samaria, 22.
Philemon, 27. Sanderson, M. G., 160, 208, 209, 210, 212,
Pierret, P., 16 n. PUIGIsAPNG VII), BOLL, OEM.
Pijnappel, J., 84. Sarat Chandra Das, 97, 147.
Pilgrim’s Castle, 44. Savenkove le Vier gauieilirs
Platona4e2s 12042774194" Scarborough Castle, 44.
Plautus, 29. Schapura, 212.
Play, 236. Schlagintweit, 146.
Plitschke, K., 68, 108, 109. Schmeller, 46.
Plowden, W. C., 194, 206. Schultze, L., 209.
Points, use of, 5, 6. Schweinfurth, G., 186, 194, 200.
Pollux, 25, 26, 27; 28: Scott, Reva Ds Gaara:
Poole, W. G., 218. Sejarah Malaya, 94.
Promotion, 9. Selenus, Gustavus, 77, 84, 87.
Pur: origin of name, 119 n. Seneca, 29 n.
Pusat, 68, 88. Seymour, R., 127.
Sha‘bi, ash-, 114.
Qabini, al-, 205. Shackell, R.S:, 213, 217, 218.
Qamiis, 37, 43 n. Shakespear, J., 169.
Quercetano, D., 77. Shamba Bolougongo, 206.
Qur'an, 115. Sharaf, ibn, 74.
Qustul, 30. Shilhak-In-Shushinak, 22.
Quyunjiq, 22. Shirreff, A. G., 47, 66, 71, 168, 213.
Shirwood, Bishop J., 84, 86.
Rabelais, 3, 76, 82, I19. Shway Yoe (Sir G. G. Scott), 111,-138.
Rattless Sir liS:108) OL L1Os 175: Sibotos, Ct. Neuenberg, 117.
Rafi‘i, ar-, 205. Sieber, J., 193:
Ragozin, 237. Simayosi Anka, Ioo.
Ramesis I, 18; III, 14. Singleton, 41.
Ra-sheps (Rashepses), 13. Sir Ferumbras, 75.
Rattray, R. S., 186. Skeat, W. W., 54, III, 116, 174.
Redstone, V. B., 102. Smith, E. W., 210, 212.
Reifferscheid, A., 25 n. Smith, H. A., §8.
Repiquet, J., 224. SHV Is 1B, GS.
Replacement capture, Io, 83. Smithy Vie Acaesiit:
Reverse holes (mancala IV games), 214. Snouck-Hurgronge, C., 68, 107, 110, 174.
Richborough, 33. Sdderbom, G., 111.
Richter, W., 26. Soham, 41.
Rigs-pula, 60. Sophocles, 26, 28.
Rig-Veda, 237. Sophron, 28 n.
Rishbeth, Mrs. (Kathleen Haddon), 238. Sparsholt, 44.
Robinson, H. C., 81. Speith, J., 181.
 GENERAL INDEX 267
Stevens, Joan, 118. VEBei. ot:
Strutt, J., 39 n., 103, 123, 142, 143. Valls, L., 78.
Suetonius, 25, 31. Vanoverbergh, W., 176.
Susay22: Varro, 33.
Sutherland, Miss L. S., 28. VELA, 89.
Svodni Kormch, 33 n. VERKEER: Origin of name, 120 n.
Veth, 110.
TABLES: origin of name, I17. Vetula, 3, 46, 84, 95.
Tacitus, 237. Vieille, La, 3 n., 38, 66.
TAEL: origin and meaning of the name, Volo-spa, 60.
56. Vopiscus, 33.
Taimiya, ibn, 205. Voth, Rev. H. R., 150.
Talbot, P. A., 185.
Talmud, 44, 114. Wall-paintings (Egypt), 13.
Taplow, 58. Walpole, Horace, 142.
Maylor, aE -72- Wappenwyll family, 66.
Tell Beit Mirsim, 20, 22. Warrington, 60.
Tell Zakariya, 22. Weber, A., 234.
Tessmann, G., 184. Weeks, J. H., 201.
Thebes, 15, 16. Werner, O., 212 n.
Theocritus, 28 n. Westermann, D., 181 n.
Thévenot, Jean de, 165. Westminster Abbey, 41.
‘THREE-IN-A-ROW GAMES, 37. White, A. C., 40.
TIGER-GAMES, 107-12. Wiedermann, A., I4.
Times, The, 60, 103. Wilkinson, R. J., 52, 107, I10, 116, 174.
Timgad, 30. Wilson, Tom, 105.
Torday, E., 206, 224. Wimose, $58, 61.
Torquemada, A., 78. Winlock, H. R., 16.
T’oung Pao, 116. Withdrawal capture, 11, 88.
Tremearne, Major A. J. N., 187. Wolfe, J., 142.
Tressan, L., 105. Wood-Martin, W. G., 41.
Treves, Sir F., 199. Woodperry, 60.
Tsuboi, Professor, 39, 54, 82, IOI, 116, 132, Woolley, Sir L., 19, 20 n., 21.
146. Worksop, 44, 142.
Tughanshah, 115.
Tutankhamen, 18. Yeung luk sz’, ror.
Twiss, R., 81. Yorkshire Weekly Post, 46.
UM reyR, SE IE, Lh, aes wepe
Zeno, 32.
Ur, 7, 19. Zwickmiihle, 45, 49.
a
se
i
no