Preface

Note on the Transliteration of Sanskrit, Persian, and Arabic Words

Explanation of the Chess Notation Used in this Work

Contractions

Books and Articles Consulted for the History of Chess

Part I. Chess in Asia

Chapter I. Introductory

Chapter II. Chess in India. I

Chapter III. Chess in India. II

Chapter IV. Chess in India. III

Chapter V. Chess in the Malay Lands

Chapter VI. Chess in Further India

Chapter VII. Chess in China, Corea, and Japan

Chapter VIII. Chess in Persia Under the Sāsānians

Chapter IX. Chess in the Eastern Empire

Chapter X. The Arabic and Persian Literature of Chess

Chapter XI. Chess Under Islam

Chapter XII. The Invention of Chess in Muslim Legend.

Chapter XIII. The Game of Shaṭranj: Its Theory and Practice. I

Chapter XIV. The Game of Shaṭranj: Its Theory and Practice. II

Chapter XV. The Game of Shaṭranj: Its Theory and Practice. III

Chapter XVI. Games Derived from Muslim and Indian Chess

Chapter XVII. The Modern Games of Islam

Chapter XVIII. Chess in Central and Northern Asia, and in Russia

Part II. Chess in Europe

Chapter I. Chess in Western Christendom: Its Origin and Beginnings

Chapter II. Chess in the Middle Ages

Chapter III. The Mediaeval Game

Chapter IV. The Early Didactic Literature.

Chapter V. The Moralities

Chapter VI. The Mediaeval Problem. I

Chapter VII. The Mediaeval Problem. II

Chapter VIII. The Mediaeval Problem. III

Chapter IX. Chess in Mediaeval Literature

Chapter X. Chessboards and Chessmen

Chapter XI. The Beginnings of Modern Chess

Chapter XII. From Lopez to Greco

Chapter XIII. From Greco to Stamma

Chapter XIV. Philidor and the Modenese Masters

Chapter XV. The Nineteenth Century

Note on the Transliteration of Sanskrit, Persian, and Arabic Words

Table of Contents

The aim of this work is threefold: to present as complete a record as is possible of the varieties of chess which exist or have existed in different parts of the world; to investigate the ultimate origin of these games and the circumstances of the invention of chess; and to trace the development of the modern European game from the first appearance of its ancestor, the Indian chaturanga, in the beginning of the seventh century of our era. The subject accordingly falls naturally into two parts: the history and record of the Asiatic varieties of chess, and the history of chess in Europe with its influence on European life and literature.

Many books have been written upon the history of chess, but none covers exactly the same field as this work. The English writers, Hyde (1694) and Forbes (1860), in the main confine their attention to Oriental chess; the great German writer, Von der Lasa (1897), treats almost exclusively of the European game. Van der Linde alone deals with both Oriental and European chess in approximately equal detail, but it is in three distinct works (1874–81).

In his great work, the Geschichte und Litteratur des Schachspiels (1874), v. d. Linde was able to incorporate the results of Professor A. Weber’s examination of the early references to chess in Sanskrit literature, and to show that Forbes’s History was both inaccurate and misleading. Since the publication of the Geschichte, however, there have been many additions to our knowledge of special features of chess history. The earliest of these were incorporated in v. d. Linde’s last work, the Quellenstudien (1881), but the later additions can only be found in isolated papers, such as those of Mr. H. F. W. Holt (Chinese chess), Herr A. v. Oefele (Malay chess), Professor A. A. Macdonell (early Indian chess), M. E. V. Savenkof (Siberian and Russian chess), Herr F. Strohmeyer (chess in mediaeval French literature), and Mr. W. H. Wilkinson (Chinese and Corean chess). It was with the idea of making all this information easily accessible to English readers that I formed the plan of writing the present work more than thirteen years ago.

To all these writers, and many others whose names will be found in the list of works consulted, I am greatly indebted, and in particular to Hyde, to v.d. Lasa (whose kindly encouragement to me in 1897 to proceed with work on the history of chess I recall with pleasure), and to v. d. Linde. But the greater part of the book is based upon my own work at original sources, especially at unpublished Arabic and early European manuscripts on chess. It was my good fortune, at an early stage of my work, to enlist the interest of Mr. John G. White, of Cleveland, Ohio, U.S.A., the owner of the largest chess library in the world. Mr. White’s generous and unfailing courtesy in placing his library freely at the service of any student of chess has been acknowledged over and over again. To me he has given not only this, but far greater help. He has repeatedly obtained copies of manuscripts which it was important that I should see, but which were inaccessible to me, and has placed these copies unreservedly at my service. Whatever in the way of completeness I have been able to achieve is entirely due to Mr. White’s help. Without that help, the book would never have been written. I must also record my indebtedness to Mr. J. W. Rimington Wilson, of Bromhead Hall, Yorkshire, who has lent me many rare books and manuscripts from the chess library which was collected by his father, the late Mr. F. W. Rimington Wilson; to Mr. J. A. Leon, who lent me the valuable sixteenth-century problem manuscript in his possession; to Mr. Bernard Quaritch, who allowed me to examine the Fountaine MS. when it passed through his hands in 1902; and to Mr. H. Guppy, of the John Rylands Library, Manchester, who made special arrangements in 1903, by which I was enabled to consult two important Arabic manuscripts at that time in the possession of the late Mrs. Rylands.

But apart from this assistance in making the original sources available, the very width of the distribution of chess and the many languages in which the literature of the game is written, would have made my task an impossible one if I had not received the help of many scholars. Among these are my father, Sir James A. H. Murray, who has not only helped me with advice of the greatest value, but has introduced me to many scholars whom otherwise I should have scarcely ventured to approach; Dr. A. C. Haddon, F.R.S.; Professor E. J. Rapson, and Dr. W. H. D. Rouse, who have helped me with Sanskrit references; Mr. S. F. Blumhardt, who translated a small Hindustani work on chess for me; Mr. E. J. Colston, I.C.S., to whom I owe the first complete account of Burmese chess; Professor D. S. Margoliouth, to whom I have taken all my difficulties in reading my Arabic sources; Bodley’s Librarian, Mr. Falconer Madan, who has dated many manuscripts for me; my sister, Miss Murray, of the Royal Holloway College, who has helped me with Icelandic references; Mr. W. W. Skeat, who has helped me in connexion with Malay chess; Mr. I. Abrahams, whom I have consulted about Jewish allusions; Mr. B. G. Laws, who has helped me to establish the European source of the problems in modern Indian textbooks of chess; and Mr. Charles Platt, of Harrow, who has allowed me to include illustrations of Oriental chessmen from his unique collection. To all these and others I express my most grateful thanks for their help. Unhappily, my thanks can no longer reach the late Professor W. R. Morfill, who gave me most valuable assistance with Russian and Czech, and the late Mr. J. T. Platts and Lieut.-Col. Sherlock, who gave me similar help with Persian and Hindustani.

In conclusion, I should like to express my personal gratification that this book is appearing from the same University Press which, more than two hundred years ago, published the pioneer work on its subject, Thomas Hyde’s Mandragorias seu Historia Shahiludii.

H. J. R. MURRAY.

CAMBRIDGE, 1913.

Table of Contents

I have departed in some particulars from the system almost unanimously adopted by Sanskrit and Arabic scholars, with a view to avoiding symbols which would probably confuse the ordinary reader. All these Oriental words will be pronounced with reasonable accuracy if the consonants are given their ordinary English pronunciation, and if the vowels are pronounced as in Italian. The following digraphs represent single sounds:—ch, dh, gh, kh, sh and th.

ch is to be pronounced as in church.

dh in Arabic words as th in this, or as z.

gh is a guttural, heavier than the Scotch ch in loch.

kh is to be pronounced as the Scotch ch in loch.

When these combinations are not digraphs, a · is placed between the two letters, as in rat·ha (to be pronounced răt-ha, not rath-a) and Is·haq (to be pronounced Is-haq, not Ish-aq). In Arabic words’ is used for the hamza (produced by a compression of the upper part of the windpipe, and practically the French h aspirée), ‘for the guttural ‘ain (produced in Arabic by a more violent compression of the windpipe, and voiced, but in Egypt and Persia practically equivalent to the hamza), and q for the deeper k which approximates to g as in gay.

Certain consonants are written with diacritical marks in order to enable the Arabic scholar to restore the written word.¹

The vowels e and o in Skr. words are always long.

Table of Contents

It has been necessary to adopt some simple method of describing the squares of the board and of recording the moves of a game which could be used uniformly for all the varieties of chess included in this work. Since the ordinary English descriptive notation does not lend itself to such adaptation, I have adopted the literal or algebraical notation which is used in all German chess books. The diagram will make clear the method of this notation, and it can obviously be extended without difficulty to a board of any size. In the cases of the Chinese and Corean games, in which the pieces are placed on the intersections of the lines dividing the board and not on the squares, a similar notation is adopted, but now the successive vertical lines are designated by letters and the horizontal lines by numerals.

Table of Contents

Table of Contents

European chess of Indian ancestry.—Asiatic games of similar ancestry.—Classification of Board-games.—Indian Board-games.—The Ashṭāpada.—Speculations on the nature of the original Indian chess.—Previous theories as to the ancestry of the game.

Historically chess must be classed as a game of war. 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 side. The players have no assistance other than that afforded by their own reasoning faculties, and the victory usually falls to the one whose strategical imagination is the greater, whose direction of his forces is the more skilful, whose ability to foresee positions is the more developed.

To-day, chess as we know it is played by every Western people, and in every land to which Western civilization or colonization has extended. The game possesses a literature which in contents probably exceeds that of all other games combined.¹ Its idioms and technicalities have passed into the ordinary language of everyday life.² The principles and possibilities of the game have been studied for four centuries, and the serious student of chess starts now with the advantage of a rich inheritance of recorded wisdom and experience. Master-play reaches a high standard, and has rightly earned a reputation for difficulty. This reputation has often been extended to the game itself, and has deterred many from learning it. Moreover, Western civilization has evolved other games, and teems with other interests for leisure moments, so that chess to-day can only be regarded as the game of the minority of the Western world. In the Middle Ages chess was far more widely played, and the precedence among indoor games that is still accorded by courtesy to it is a survival from the period when chess was the most popular game of the leisured classes of Europe.

The ancestry of this European chess can easily be established. A number of the mediaeval European chess terms can be traced back by way of Arabic to Middle Persian. Thus we have

The name of the game in most of the European languages, e.g. Eng. chess, Fr. échecs, It. scacchi, can be traced back, through the Latin plural scaci (scachi, scacci, meaning chessmen), to the Arabic and Persian name of the chess King, shāh.

The names of the other chessmen—King and Pawn (L. pedo, a foot-soldier), everywhere; Horse, in Southern Europe—reproduce the meaning of the names of the corresponding men in the Arabic and Persian games.

The names of the game of chess in modern Spanish or Castilian (ajedrez) and Portuguese (xadrez) not only confirm this evidence, but supplement it by taking the pedigree a step farther back. For these two forms appear in older Castilian as acedrex, and this word is simply the Arabic ash-shaṭranj, the shaṭranj, in a European dress. Shaṭranj, again, is only an Arabicized form of the Middle Persian chatrang, and this Persian word is an adaptation of the Sanskrit chaturanga. All these terms are in their respective languages the ordinary names for the game of chess.

The names of the chessmen in Persian and Sanskrit are synonymous. In each game there was a King, a Counsellor, two Elephants, two Horse, two Chariots, and eight Foot-soldiers.

This philological evidence derives some support from the documentary evidence. The earliest works which make mention of chess date from about the beginning of the 7th century A.D., and are associated with N.W. India, Persia, and Islam. It is difficult to assign exact dates, but the oldest of a number of nearly contemporary references is generally assumed to be a mention of chess in a Middle Persian romance—the Kārnāmak—which is ascribed with some hesitation to the reign of Khusraw II Parwīz, the Sāsānian king of Persia, 590–628 A.D. The others belong to N.W. India.

It is interesting to note that early Persian and Arabic tradition is unanimous in ascribing the game of chess to India. The details naturally vary in different works, and the names in the tradition are manifestly apocryphal. Chess is usually associated with the decimal numerals as an Indian invention, and its introduction into Persia is persistently connected with the introduction of the book Kalīla wa Dimna (the Fables of Pilpay) in the reign of the Sāsānian monarch Khusraw I Nūshīrwān, 531–78 A.D., and European scholars of Sanskrit and Persian generally accept the traditional date of the introduction of this book as established. The so-called Arabic numerals are well known to be really Indian.

Finally, a comparison of the arrangement and method of the European game of the 11th to 13th centuries A.D. with the Indian game as existing to-day and as described in the earlier records supports the same conclusion. In both games the major pieces occupy opposite edges of the board of 8 × 8 squares, and the Foot-soldiers are arranged on the row in front of the major pieces. The corner squares (a1, a8, h1, h8) are occupied by the Chariol with identical move in most of the games;⁴ the next squares (b1, b8, g1, g8) by the Horse with the well-known move of the Knight; the third squares from the corners (c1, c8, f1, f8) by the Elephant;⁵ and the two central squares (e1, e8, d1, d8) by the King and Counsellor respectively with moves that were for long the same in India, Persia, Islam, and Europe.⁶ The move of the Foot-soldiers, arranged on the 2nd and 7th rows, was also for long the same in the chess of all these countries.

We must accordingly conclude that our European chess is a direct descendant of an Indian game played in the 7th century with substantially the same arrangement and method as in Europe five centuries later, the game having been adopted first by the Persians, then handed on by the Persians to the Muslim world, and finally borrowed from Islam by Christian Europe.

Games of a similar nature exist to-day in other parts of Asia than India. The Burmese sittuyin, the Siamese makruk, the Annamese chhôeu trâng, the Malay chātor, the Tibetan chandaraki, the Mongol shatara, the Chinese siang h‘i, the Corean tjyang keui, and the Japanese sho-gi, are all war-games exhibiting the same great diversity of piece which is the most distinctive feature of chess.

There is naturally far less direct evidence respecting the ancestry of these games than in the case of European chess, but there can be no doubt that all these games are equally descended from the same original Indian game. The names sittuyin (Burmese), chhôeu trâng (Annamese), and chandaraki (Tibetan) certainly, and the names chātor (Malay) and shatara (Mongol) probably, reproduce the Sanskrit chaturanga. The names of some of the pieces in the Malay, the Burmese, and probably the Siamese games, have been borrowed from the Sanskrit.

If we examine the nomenclature of these games we also find the same meanings recurring throughout. Thus we have—

The Malay, Tibetan, and Mongol games are played on a board of 8 × 8 squares, and the initial arrangement of the pieces corresponds closely to the Indian game. The three games of Further India are played on a board of the same size, but the arrangement of the pieces differs from that of the Indian game. The moves of the chessmen are consistent with an Indian ancestry.

The relationship of the Chinese, Corean, and Japanese games is not so obvious. The first two are played on the lines, and not on the squares, of a board of 8 × 8 squares with a space between the 4th and 5th rows which virtually makes the board one of 8 × 9 squares; the third is played on the squares of a board of 9 × 9 squares. There is, however, no doubt that both the Corean and the Japanese games are derivatives of an older form of the Chinese game. Chinese works refer to the introduction of modifications in their game after 1279. These games introduce new pieces, but the salient fact remains that the Chariot with the move of the Rook (modified in Japan) occupies the corner squares (a1, &c.), and the Horse with the characteristic move of the Knight (slightly modified) occupies the adjoining squares (b1, &c.). This coincidence is too striking to be dismissed as merely accidental. Moreover, it is well known that other Chinese games are of Indian origin.

We may contrast the position of these games in Asia with that of chess in Europe. If we except Japan, there are only the beginnings of a literature. Each generation accordingly has to start again from the commencement and to evolve its own science of the game. The standard of play remains of necessity low, and there is nothing to deter any one from learning to play. The game has few rivals with which it must compete for popular favour, and it has had no difficulty in most places⁷ in retaining the first place. Thus the majority of Asiatics are chess-players, and chess may without exaggeration be described as the national game of Asia.

It is in the wider sense, in which I have just used the word, that I propose to use chess in this book. I include under it all the games which I trace back to the Indian chaturanga, and all the freak modifications that have been attempted from time to time. The first part of this history is devoted to a record of the Asiatic varieties of chess, and the evidence rapidly summarized above will be developed at greater length in the sequel. The broad lines of the diffusion of chess from India are fairly clear. Its earliest advance was probably westwards to Persia; the eastward advance appears to have been rather later, and at least three lines of advance may be traced. One route took the game by Kashmīr to China, Corea, and Japan. A second, possibly the same route by which Buddhism travelled, took chess to Further India. At a later date chess spread from the S.E. coast of India to the Malays. The route by which the game reached Tibet and the Northern tribes of Asia is still doubtful. Persia had meanwhile passed on chess to the Eastern Roman Empire, and, as a result of the Muhammadan conquest of Persia, Islam learnt the game. Henceforward the Muslims became the great pioneers of chess, carrying their game as far west as Spain, and east to India where they imposed the Arabic nomenclature on the Northern and Central Provinces of the Peninsula. Christian Europe had begun to learn chess from the Moors as early as 1000 A.D. From the Mediterranean shores it spread northwards over France and Germany to Britain, to the Scandinavian lands, and Iceland.

In its outward furniture chess is only one of many games which require a specially arranged surface for play. Games of this type are conveniently grouped under the generic name of Board-games, Ger. Brettspiele, although, as Groos⁸ has pointed out, the name is not a very fortunate one, since the surface of play is not always a board. Board-games are not only of very wide distribution to-day, but are also of great antiquity. They are by no means confined to the more civilized races: with the exception of the native tribes of Australia and New Guinea, practically every known people has its game or games of this type. It has also been remarked that the difficulty of a board-game is no criterion of the development of the race playing it, for some of the most involved and complicated varieties known are played by tribes that stand lowest in the scale of civilization. Board-games were played by the early inhabitants of Egypt; boards and pieces have been found in tombs even as old as the pre-dynastic period (a. 4000 B.C.),⁹ they are depicted in paintings in tombs of the Fifth Dynasty (3600–3400 B.C.),¹⁰ and the masons who built the temple at Kurna (1400–1333 B.C.) cut boards on slabs which were afterwards built into the roof of the temple.¹¹ Boards, apparently for games, have been found in prehistoric ruins in Palestine.¹² Board-games are mentioned in the earliest Buddhist literature of India,¹³ and in early Chinese works.¹⁴ They were played in classical times in Greece and Rome,¹⁵ by the Celts in Ireland and Wales before the Norman Conquest of England, by the Norse vikings before they began to harry the coasts of England and France,¹⁶ and by the native tribes of America before the time of Columbus.¹⁷

All known board-games, greatly as they vary in arrangement and method of play, appear to fall into one or other of three well-defined groups:

(1) Race games, in which the men are moved along a definite track. The typical European example is the game of Backgammon (tables, nard).

(2) Hunt or Siege games, in which one side endeavours to block or confine the adversary. The typical European example is the game of Fox and Geese.

(3) War games, in which the capture of prisoners plays a considerable part. The typical European example is the game of Chess.

This classification is convenient, but it must not be pushed too far. In particular, it must not be assumed without further inquiry that it involves any necessary connexion between the individual games of different groups, or even of a single group. However tempting it may be to assume a common ancestry for board-games, it is clear from a closer examination of the various methods of play that the majority have arisen independently, and that only in the case of a small minority in any class is there any evidence of a common origin. The sameness of type which is the foundation of the above classification is at most due to the fact that the games are ‘based upon certain fundamental conceptions of the universe’ (Culin, Korean Games), but more probably, in my opinion, to the universality of the activities which the games symbolize.¹⁸ Identity of origin can only be established by the evidence of reliable historical documents, by the linguistic evidence derived from the nomenclature of the games, or by the fact that these show so great an identity of feature that the chances of independent invention are mathematically infinitesimal.

The existing games which I include under the name of chess form one of the few groups of games whose common ancestry can be established in this way. It will obviously be far more difficult to carry the pedigree farther back, and to discover the origin or relationships of the parent Indian chaturanga, a game already in existence in the 7th century of our era, in still older games. We shall first have to ascertain what board-games were in existence in India at that remote period, and to attempt to elucidate their nature.

Unfortunately, the general characteristics of early Indian literature are not very favourable for such an inquiry. The earlier Sanskrit literature of the Vedic age, and also of the later centuries when the Brahmanas and Sutras came into existence, was religious in tone and almost entirely poetical in form, and references to games must be exceptional. The later Sanskrit literature gradually extended its field to include secular subjects in general, but as it widened its field the defects of its literary style became more pronounced, and the conceits of the poetry and the extraordinarily condensed character of the prose deprive the allusions of definiteness, and leave too much to depend on the view of the commentator or the personal fancy of the translator. Our knowledge of the older Indian games is thus very vague, and based only upon the comparison of passages, all more or less obscure.

But we do know that board-games were in existence in N.W. India and the Ganges valley considerably before the commencement of the Christian era. We know this from the occurrence in Sanskrit works of words which are used as the names of boards or surfaces upon which games were played. The commonest of these words is phalaka, but this is simply a generic term for a game-board and conveys no information as regards shape, size, or arrangement. There are next the terms used in connexion with the simplest forms of dice-play, in which everything turns upon the result of throwing the dice and nothing in the nature of a game with pieces is required. Obviously, all that is necessary in this case is a level surface upon which the dice may fall, and Lüders (Das Würfelspiel im alten Indien, Berlin, 1907, 11–15) has shown that adhidevana (used in the Atharva Veda, and usually translated dice-board) meant simply a smooth flat surface excavated in the ground for this purpose. Of more importance for our present purpose is a group of terms which are restricted to boards of definite shape and arrangement. There are two words of this kind: ashṭāpada, meaning a square board of 64 squares, 8 rows of 8 squares, and dasapada, meaning a similar board of 100 squares, 10 rows of 10 squares. These boards were employed for a more complicated form of game in which the use of the dice was combined with a game upon a board (Lüders, op. cit., 65). Both terms appear to have been used also for the games played upon these boards.

The ashṭāpada would seem accordingly to have been identical in shape with our chessboard or draughtboard, and so it is often translated, though the rendering is to be deprecated as suggesting to the ordinary reader that the board was used for a rudimentary form of one of these games. For draughts there is no evidence at all, for chess none before the 7th c. A.D. Still, the coincidence is so striking that it is worth while to try to discover what the ashṭāpada game really was, in order to see whether it has not some connexion with the rise of chess.

The meaning of the word is established by Patañjali in his great commentary on the grammar of Pāṇini, the Mahābhāshya, which, according to Macdonell (Skr. Lit., 431), was written between the latter half of the 2nd c. B.C. and the beginning of the Christian era. It is here¹⁹ defined as ‘a board in which each line has 8 squares’. In the absence of any reference to any alternate colouring or chequering of the squares, we may assume that it was unchequered, like all other native Asiatic game-boards. Two early comparisons suggest that the ashṭāpada was a familiar object. In the first book of the Rāmāyaṇa,²⁰ according to Jacobi added after the 2nd c. B.C., the city of Ayodhyā (Oudh) is spoken of as ‘charming by reason of pictures consisting of ashṭāpada squares, as it were painted’. The regular plan of the city is probably intended, and the passage may be compared with later ones from Muslim historians. Thus Ḥamza al-Iṣfahānī (c. 300/912), writing of the building of Jundī Shāpūr by the Sāsānian king Shāhpūr (240–270 A.D.), says: ‘the plan of this city was after the fashion of a chessboard; it was intersected by 8 times 8 streets,’ to which a later Persian historian adds the pertinent comment, ‘the figure was after this fashion, but chess had not yet been invented at that time.’ The later geographer Mustawfī (740/1340)²¹ has a similar statement about the plan of Nīshāpūr in Khurāsān: ‘In the days of the Chosroes, as it was reported, the old town of Naysābūr had been originally laid out on the plan of a chessboard with 8 squares to each side,’ There is also a passage in a Northern Buddhist work, cited by Burnouf in his Lotus de la bonne loi, Paris, 1852–4, 383, in which the world is described as ‘the earth on which ashṭāpadas were fastened with cords of gold’—probably alluding to the division by roads, seas, and mountains, or to the succession of field, forest, and desert.²²

Of more importance is a passage in the Pali²³ Brahma-jāla Sutta, or Dialogues of the Buddha,²⁴ according to Rhys Davids one of the earliest of Buddhist documents, purporting to record the actual words of Gotama himself, and dating back to the 5th c. B.C. The Buddha is contrasting the conversation and thoughts of the unconverted man with those of the disciple:

It (sect. 7, p. 3) is in respect only of trifling things, of matters of little value, of mere morality, that an unconverted man when praising the Tathāgata, would speak. And what are such trifling, minor details of mere morality that he would praise?

He then proceeds to enumerate all the many trifles which occupy the thoughts of the unconverted man, and finally comes to games, and gives us a most interesting and valuable list of games—quite the oldest known—which from its interest I quote entire:

Or (sect. 14, p. 9) he might say, ‘Whereas some recluses and Brahmans while living on food provided by the faithful continue addicted to games and recreations; i.e. to say—

1. Games on boards with boards with 8 or 10 rows of squares.

2. The same games played by imagining such boards in the air (Pāli, ākāsaṃ).

3. Keeping going over diagrams drawn on the ground, so that one steps only where one ought to go.

4. Either removing the pieces or men from a heap with one’s nail, or putting them in a heap, in each case without shaking it. He who shakes the heap loses.

5. Throwing dice (Pāli, khalikā).

6. Hitting a short stick with a long one.

7. Dipping the hand with the fingers stretched out in lac, or red dye, or flour water, and striking the wet hand on the ground, or on a wall, calling out ‘What shall it be?’ and showing the form required—elephants, horses, &c.

8 Games with balls (Pāli, akkhaṃ).

9. Blowing through toy pipes made of leaves.

10. Ploughing with toy ploughs.

11. Turning somersaults.

12. Playing with toy windmills made of palm leaves.

13. Playing with toy measures made of palm leaves.

14, 15. Playing with toy carts, or toy bows.

16. Guessing at letters traced in the air, or on a playfellow’s back.

17. Guessing the playfellow’s thoughts.

18. Mimicking of deformities.

Gotama the recluse holds aloof from such games and recreations.’

This passage is quoted at length in many other early Buddhist works, e.g. in Vinaya, ii. 10, and iii. 180. The translation naturally depends considerably on early native commentaries, and it must be remembered that the earliest commentators are considerably later than the original; indeed they only appeared when changes in the spoken language made the written work archaic and unintelligible to the ordinary reader. The commentator was often in a worse position than the modern scholar for interpreting the text, and we often find his explanation absurd or impossible. We are accordingly compelled to accept the above translation with some reserve.²⁵

We are only concerned now with the first two of the games named. These are the ashṭāpada—here in its Pali form aṭṭhapada—and the dasapada. One of the two commentators used by Rhys Davids, the Sinhalese Sanna, who belongs to the 10th C. A.D. or even later, says that each of these games was played with dice and pieces (poru, from purisa = men), such as Kings and so on.²⁶ His evidence is far too late to be of any value as to the nature of the games in question, but is important as showing that these boards were still used for dice games in his day in Ceylon. Yet, if the second sentence is accurately translated, the games must have been of a character which permitted ‘blindfold’ play without the use of material boards.

The game on the ashṭāpada also falls into condemnation in an early Brahman work, the Sutrakrilānga.²⁷ The devout Brahman, we are told,

should not learn to play ashṭāpada, he should not speak anything forbidden by the law, a wise man should abstain from fights and quarrels.

A more illuminating reference is to be found in the Harivaṃsa, or Family of Vishṇu, a supplementary book to the Mahābhārata, and generally recognized as a later addition. Macdonell (Skr. Lit., 287) has, however, shown that the Mahābhārata, including the Harivaṃsa, must have attained to its present form by at least 500 A.D. The passage²⁸ recounts a meeting for dice-play between Rūkmin and Balarāma. The former had the reputation of being an expert at dice, the latter was fond of it, but not very skilled in play. Enormous stakes were laid, and Rūkmin won thrice in succession. Finally, sorely provoked by Rūkmin’s expressions of triumph, Balarāma exclaimed, ‘Prince, I wager the vast sum of 100,000 millions, do you accept it? Let us throw the black and red dice on this splendid ashṭāpada.’ Rūkmin made no reply, but threw and lost. Then only did Rūkmin reply, ‘I refuse the wager.’ Neither this, nor Rūkmin’s continued references to his victory, upset Balarāma’s self-control, but when a voice from the skies awarded the victory to him on the ground that ‘silence gives consent’, Balarāma’s long-restrained wrath blazed forth, and seizing the large golden ashṭāpada, he struck Rūkmin to the ground. A second blow broke the teeth of the King of Kalinga. Then, tearing up one of the golden pillars of the hall, Balarāma strode forth, wielding it as a club.²⁹

We may probably find in this story a reason for the condemnation which Buddhist and Brahman alike pronounce upon the game ashṭāpada. Neither religion countenanced dicing, but neither has been able at any time to suppress it in India. Too great stress has been placed upon the efficacy of legislation, such as is to be found in the Code of Manu, against the use of the dice.³⁰ It is abundantly evident from the whole extent of Sanskrit literature that gambling with dice has been at all times the chief recreation in India. One of the very few secular poems in the Rigveda, occurring in the very oldest part of the collection, which can hardly be put later than 1000 B.C., contains the lament of a gambler who is unable to tear himself away from the dice, although he is fully conscious of the ruin he is bringing upon himself and his home. Lüders (op. cit.) has collected a large number of instances from the epic literature which show the extent of the passion for dicing in post-vedic times. In the Mahābhārata, Nala and Yudhishthira are represented as gambling away their very kingdoms in dice-play.³¹ The Arabic historian al-Maṣ‘ūdī, writing about 950 A.D., draws a lurid picture of what was currently believed in his day of the gambling propensities of the Indians. He is writing of the uses of ivory, and continues:³²

But by far the most frequent use of ivory is for the manufacture of men for chess and nard. Several of the chessmen are figures of men or animals, a span high and big, or even more. During the game a man stands by, specially to carry the men from one square to the other. When the Indians play at chess or nard, they wager stuffs or precious stones. But it sometimes happens that a player, after losing all his possessions, will wager one of his limbs. For this they set beside the players a small copper vessel over a wood fire, in which is boiled a reddish ointment peculiar to the country, which has the property of healing wounds and stanching the flow of blood. If the man who wagered one of his fingers loses, he cuts off the finger with a dagger, and then plunges his hand in the ointment and cauterizes the wound. Then he returns to the game. If the luck is against him he sacrifices another finger, and sometimes a man who continues to lose will cut off in succession all his fingers, his hand, his fore-arm, his elbow, and other parts of his body. After each amputation he cauterizes the wound with the ointment, which is a curious mixture of ingredients and drugs peculiar to India, of extraordinary effectiveness. The custom of which I have spoken is a notorious fact.

At the present day games of chance are among the most popular of Indian games, and are associated with religious festivals, especially with those in which it is necessary to keep watch the whole night through.³³

The ashṭāpada is also mentioned in an account of a game between Sakuni and Yudhishthira in Amarachandra’s Bālabhārata (II. v. 10 ff). In this game two dice (respectively red and black) are used, and each player has an ashṭāpada upon which he throws his die.³⁴ The game was played with pieces (sāri), of which half were red and the other half were black. These are moved in obedience to the throws of the dice; the ‘clatter’ which they make when placed upon the new position is mentioned, and the sāri are compared to monarchs, since like these they are set up, moved, taken captive, and released.

It seems clear that we have to do here with a game of the race-game class. We may find some confirmation for this conclusion from the comparative study of other Asiatic board-games in which dice are used to define the movements of the men. In India itself there exist a number of examples of games of this class, of which the best known are the games pachīsī and chaupur, which are played upon a four-armed board.

Games of this type appear to have been practised over the greater part of the world from the earliest times. A wide selection of examples is to be found in Mr. Stewart Culin’s books on games.³⁵ The underlying principle is practically the same in all. The board is arranged so that the divisions or points constitute a track along which the men (in Asia commonly called horses or dogs) are moved in obedience to the throws of the dice or equivalent implements (e.g. staves, shells, seeds, teetotums). The players, who may be two or more in number, are each given a certain number of men whom they have to enter on, move through, and remove from the board in a prescribed manner. Any player can remove, with certain limitations, an opponent’s man from the board by playing one of his own men to the point occupied by the former, and the man so removed has to commence again from the beginning. The player who first succeeds in removing all his men from the board after completing his appointed track, wins the game.

Probably the oldest and simplest Asiatic game of this type is the game for two players which we call backgammon. It is now played with little variety over all Southern Asia, from Syria to Japan. Chinese records mention its introduction from India with the name t‘shu p‘u (= Skr. chatush-pada, mod. Indian chaupur) as early as 220–65 A.D. Weber³⁶ has collected a number of references to games of this character from early Indian literature, the earliest being from the Mahābhāshya, in a passage in which Patañjali discusses Pāṇini’s explanation of the word ayānayīna,³⁷ in which the termination -ina has the force of ‘to move to’.

Board for Pachīsī and Chaupur.³⁸

Gavalata Board (Culin, C. & P. C., 851).

Ashta Kashte Board (Falk., 265).

It was possibly the desire to frame a game for four players on similar lines which led to the invention of the four-armed and square boards of which we have several Indian examples. All these boards exhibit a further modification in the special markings that are placed on particular squares. The device is not peculiar to Indian games: it represents an obvious way of adding additional interest to the game which occurred independently to players in many regions. A man which is played to one of these cross-cut squares is treated differently from one played to an unmarked point. It may secure the option of a shorter route home, as in the Corean nyout. It may secure immunity from capture so long as it occupies that point, as in these Indian games, and indeed in the majority of Asiatic race-games. It may be penalized by being compelled to return to the starting-point again, as in the American games of this class. It may be subjected to other penalties, or be given other privileges, as in the various race or promotion games which are invented annually in Europe, America, and elsewhere.

Board, Dice, and Men used in Saturankam (chaturanga), (Parker, 695).

Sīga Board (Parker, 607), The arrows show the direction of the moves. [The same game is in the Museum für Völkerkunde, Berlin, 1. c. 5708a, as Sadurañgam.]

Although specially arranged for four players, these games can easily be adapted to use by two players only, and the Indian games of which I give diagrams are often so used. The Ceylon game Gavalata is played by two or four players. When two play, the men enter at A and B respectively, when four, the centre point on each side is the point of entry for one of the players. Each player has one or two cowries instead of men, and four or five cowries are used instead of dice. The men move in the direction of the arrows, and the object is to traverse all the squares to the centre. A player returns an adversary to the starting-point when he plays one of his men to the same point occupied by the adversary, unless it stands on a cross-cut square, or castle. Sīga, which Mr. Parker (Ancient Ceylon, London, 1909, 607) describes as played in Colombo, is the same game, but men similar to the one shown in the diagram of saturankam are used when a proper board (generally of cloth) is employed. Often, however, the game is played upon a board marked for the occasion on the ground, and then the players make use of sticks of distinctive colour or length which they set upright in the square occupied. Saturankam and Ashta kashte are similar games on boards of 81 and 49 squares respectively. A similar game is probably depicted in the gambling scene Chitupada Sila on the coping of the Stupa of Bharhut, a Buddhist monument illustrative of Buddhist legend and history which is now considered to belong to the 4th c. A.D. Here we have four men squatting in pairs on opposite sides of a board of 6 × 6 squares. Beside the board lie 7 square pieces, 6 in a group and one nearer the board and in front of one of the players. They appear to be rudely engraved with dissimilar patterns, and have been variously identified as dice (or similar implements) or coins. The board is scratched on the ground and shows no cross-cut squares, but a short stick has been set up on one of the squares which—from the analogy of Sīga—probably represents a man in course of play.

The Bharhut Board.³⁹ The numbers show the positions of the players.

The existing board-games of this special type in Southern India and Ceylon are all played on boards with an odd number of squares, so that there is a single central square which serves as point of exit for all four players alike. In Pachīsī on the other hand, each player has his own point of exit, and there seems no reason why a similar arrangement should not have been tried upon a square board. In this case the square would obviously be one with an even number of points, and the four central points would serve as the four points of exit for the four players.

It is to this more complicated type of race-game that I assign the early Indian game on the ashṭāpada board. I find support for my belief in a peculiarity of the modern Indian chessboard which has no importance for chess and has never been explained in a satisfactory manner. On all native chessboards which I have seen, certain squares are cross-cut precisely as in the games of Pachīsī and Gavalata. Native books from the time of Nīlakaṇṭ·ha (17th c.) onwards carefully preserve the marked squares, but attempt no explanation of them. They have even survived the chequering of the board. In their complete form the boards contain no less than 16 cross-cut squares—a1, a4, a5, a8, d1, d4, d5, d8, e1, e4, e5, e8, h1, h4, h5, h8. Other boards omit some of these markings, but do not substitute other cross-cut squares for them. In the chequered boards the markings on the four central squares are not completed.

THE MARKINGS ON MODERN INDIAN CHESSBOARDS.

A. Hyde, ii. 74; Nīlakaṇṭ·ha; Brit. Mus.; Platt Collection.

B. Weber (v. d. Linde, i. 124, Bombay); Poona; Platt Collection.

C. Chequered board in Platt Collection.

D. Weber (v. d. Linde, i. 124, Tanjore).

E. Delhi.

F and G. Patiala.

This peculiarity is not confined to the Indian chessboard. There are markings on the Burmese, Malay, Chinese, and Corean boards, but these do not correspond to the Indian markings, and in some cases are now associated with special features of play. The older Muslim literature of chess makes no reference to the existence of marked squares, but Mr. Falkener possessed a modern Turkish chess cloth in which the squares a4, a5, d1, d8, e1, e8, h4, h5 are marked in one way and d4, d5, e4, e5 in another and more elaborate way.⁴⁰

The explanation of these cross-cut squares is, I believe, to be found in the fact that the Indian chessboard is simply the old ashṭāpada board, and preserves its original features, although their purpose has long been forgotten. The ashṭāpada game was, I believe, very similar to the modern gavalata. If two players played, each entered his men at opposite sides of the board; if four, then at each edge. The track ran round the outer edge, then round the inner blocks of 36 and 16 squares, and finished in the centre of the board. The cross-cut squares were citadels, or squares on which a man was immune from capture. As will be seen in the following chapter, this hypothesis provides a simple explanation for the curious fact that the Ceylon game of this type is now called saturankam, i.e. chaturanga.

The game of chess was invented when some Hindu devised a game of war, and, finding the ashṭāpada board convenient for his purpose, adopted it as his field of battle.⁴¹ The fact that he gave his game a new name, chaturanga, shows that his game had no connexion with the game of whose board he availed himself. The meaning of this name is perfectly plain. It is an adjective, compounded from the two words chatur, four, and anga, member, limb, with the literal meaning having four limbs, four-membered, quadripartite. In this original sense it appears in the Rigvēda (X. xcii. 11), in reference to the four-limbed human body, and in the Satapātha Brahmaṇa (XII. iii. 2. 2). It also occurs repeatedly in the Mahābhārata (which existed in its present form by 500 A.D.), in Rāmāyaṇa (which goes back in its oldest form to the 5th c. B.C.), in Kāmandaki’s Nītisāra (dating from the beginning of the Christian era), and in the Atharva Veda-Parisiṣṭas (which are not earlier than 250 A.D.), either in agreement with the word bala, army, or used absolutely as a feminine or neuter substantive, in the sense of army composed of four members, and army generally. It is clear that the word chaturanga became the regular epic name for the army at an early date in Sanskrit. Weber states that the use of the word, as also of the variant chaturañgin, is not only common in Sanskrit, but also in Pali.

What was meant by the four members of the Indian army is perfectly plain from the repeated connexion of the word chaturanga with chariots, elephants, cavalry, and infantry. In Rāmāyana (I. lxxiv. 4), in Mahābhārata (III. 1504. 4), and in Amarakoṣa (III. 8. 21), the army is expressly called hasty-ashwa-rat·ha-padātam, the total or aggregate of elephants, horses, chariots, and foot-soldiers. Macdonell (op. cit., 118) notes that this was the regular composition of the complete Indian army at least as early as the 4th c. B.C., for the Greek accounts of the invasion of N.W. India by Alexander, in 326 B.C., state that the army of Pauras consisted of 30,000 infantry, 4,000 cavalry, 200 elephants, and 300 chariots. The Greek historian Megasthenes, who spent some time at the court of Pāṭaliputra (Patna) about 300 B.C., when speaking of the military administration of the Indian state, says that there were six departments responsible for the management of the elephants, cavalry, chariots, infantry, baggage, and boats.⁴² The Code of Mann (vii. 185) also speaks of an army of six parts, to which the scholiast Kullūka Bhaṭṭa (16th–17th c.) adds that the six parts are hasty-ashwa-rat·ha-padāti-senapāti-karmakara, or elephants, horses, chariots, foot-soldiers, general, and camp-followers, i.e. the regular army with its commander and that motley following that always attends an Indian army on its march, and yet adds no fighting-strength to it on the day of battle.⁴³ The Nītisāra of Kāmandaki, ‘a work of policy dating probably from the early centuries of our era’ (Macdonell, JRAS., 118), contains an important and instructive chapter (ch. xix) of 62 slokas, which specially treats of the chaturangabala, or army. The chapter states that the army is composed of elephants, chariots, horse, and infantry; it discusses the ground most suitable for the evolutions of each of these members; it estimates a horseman as equal to three foot-soldiers, and the elephant and chariot as each equal to five horsemen. It suggests several arrangements as suitable for use in war, e.g., infantry, horse, chariots, elephants; elephants, horse, chariots, infantry; the horse in the centre, the chariots next, and the elephants on the wings.⁴⁴

We are, therefore, entitled to conclude that the fourfold division of the Indian army into chariots, cavalry, elephants, and infantry, was a fact well recognized already before the commencement of our era.⁴⁵

The same four elements—chariots, horse, elephants, foot-soldiers—appear as four out of the six different types of force in the board-game chaturanga. The remaining types prefigure individuals, not types of military force. The presence of the King needs no justification. The addition of the Minister or Vizier is in complete agreement with Oriental custom, and the Code of Manu (vii. 65) lays stress upon the dependence of the army on him. The self-consistency of the nomenclature and the exactness with which it reproduces the composition of the Indian army afford the strongest grounds for regarding chess as a conscious and deliberate attempt to represent Indian warfare in a game. That chess is a war-game is a commonplace of Indian, Muslim, and Chinese writers.

But the parallelism does not end with the name of the game and the chessmen. It extends to the termination of the play. The immediate object of warfare is the overthrow of the enemy, and in early times this object was secured with equal certainty either by the capture or death of the opposing monarch, or by the annihilation of his army. These are exactly reproduced by the two methods of winning in early chess—the checkmate and the baring of the opponent’s King.

It would be unreasonable to assume that the attempt to carry out the idea of arranging a war-game between Indian armies upon the ashṭāpada was immediately successful in producing the game as it appears in the oldest records, or even a workable game. But the comparative evidence of the Indian and non-Indian forms of chess shows that the period of experiment was practically past before the game had spread from its earliest centre, and that the moves, method of play, and rules were broadly settled as we know them in the oldest records. Still, one or two of the points of difficulty in the development of the game must be briefly considered.

1. The number of players. I have already suggested that the use of square boards for race-games may have resulted from the desire to give the track a fourfold symmetry which would allow of four players playing at one time. We have, however, seen that the ashṭāpada was frequently used by two players only, so that we cannot assume that a square board necessarily suggested a game for four players. Moreover, the race-game and the war-game are not really similar. The former is a one-dimensional game, since it only requires a track; the latter is a two-dimensional game and needs a surface.

We shall see that by the year 1000 there were Indian varieties of chess in existence both for two and for four players. In each variety the four elements of the chaturangabala are completely represented. In the two-handed game the King and his Minister are added, in the four-handed game the King only. The advocate for the priority of the four-handed chess might argue that its representation presents a closer parallel to the Indian army than does the chess for two players. He could also point to the fact that Indian policy has always had an eye on a warfare in which four kings were concerned, to wit, the aggressor, his foe, the neutral, and the one called the ‘middle-most’.⁴⁶ But I do not think that either argument carries much weight, I have already expressed the opinion that the presence of the Minister in a war-game can be justified from Sanskrit discussions of his functions. And this philosophical view of warfare as involving four Kings can only be looked upon as a generalization, for it is obvious that the aggressor and his foe would be quite capable of conducting a war without the intervention of the other two monarchs. So far as Indian evidence goes, I do not think that it is decisive for or against the priority of either form of chess, though the probabilities are stronger for the priority of the two-handed game. On the other hand the comparative evidence of the non-Indian games tells strongly in favour of the original game of chaturanga having been for two players. This conclusion seems to me also the more natural one. The development of a four-handed game may have been helped by considerations like the above: the analogy of the development of four-handed race-games from the simpler two-handed variety supplies a more probable reason for its appearance.

2. The arrangement of the forces. Kāmandaki’s treatise shows us that the Indians paid considerable attention to the theoretical arrangement of an Indian army on the battle-field. The problem how best to arrange the elements on the ashṭāpada was a far simpler one, since all disturbing factors were eliminated. The advantages of a symmetrical arrangement must have been obvious from the first, and we may explain the duplication of the chariot, horse, and elephant, and the eight foot-soldiers in this way. The larger number of the last named is explained by the fact that the infantry is numerically the largest part of the army. The positions of the King and his Minister on the two central squares of the first row, and of the Foot-soldiers on the eight squares of the second row, follow so naturally that I think they must have been so from the commencement. But there is no obvious reason why the remaining pieces should be arranged in any particular way, and the existing arrangement, a1 Chariot, b1 Horse, c1 Elephant, was probably only arrived at after experiment. The position of the Horse (b1, g1) is so invariable in all forms of chess, that it must have been fixed very early. As regards the other pieces, the earlier Indian references show that there was uncertainty until comparatively late in India, and now the Chariot, now the Elephant appears on the corner squares. The comparative evidence of the non-Indian forms of chess points, however, to the arrangement a1, Chariot; b1, Horse; c1, Elephant; d1 and e1, King and Minister; f1, Elephant; g1, Horse; h1, Chariot, as having been the more usual Indian one.

3. The powers of move. We have seen from Kāmandaki that the four elements of the Indian army were of very different values. If war was to be represented by a game, it was necessary to discover some means of reproducing this difference of value. This was cleverly achieved by the original idea of giving different moves to the chessmen, so that the freedom or range of the move should suggest roughly the actual method of movement of the original element in war. The general identity of move in the earlier forms of chess the world over shows the skill with which the idea was carried out: the variation in move of the Elephant recorded in early Indian chess, and exhibited to-day in existing Asiatic forms of chess, may be taken as showing that the final result was only obtained after experiment.

4. The method of play. All race-games are dice-games, and it is probable that all board-games were in the first instance played by means of dice or other implements of similar import. There is no reason, as far as I can see, why we should make an exception to this in the case of chess. Previous writers have approached the question with a priori arguments. V. d. Linde (i. 79–80) lays stress on the incompatibility of dice and chess, and considers it a dualism that could not be original. V. d. Lasa (1) thought that the greatest probability was in favour of the original game having been a pure game of combination. Macdonell (JRAS., 140) is disposed to take the view that there was a dice-age in the development of chess, as offering a more natural development than that which the opposite view offers. The evidence of the earlier Indian references to chess is purely negative. Dice are nowhere mentioned, but nowhere of necessity excluded from use. It is only at a comparatively late date that we begin to hear of varieties of chess in which the moves were given by the throws of the dice. The four-handed game was a dice-game in its earlier history. The Muslims played their oblong chess on a board of 4 × 16 squares with the help of dice. Even in Europe varieties of dice-chess were not unknown in the 13th c., though it is probable that some of these were of European invention.

But the later Indian references to the two-handed chess, and the comparative evidence of the non-Indian games show that at quite an early period the possibility of playing chess without dice had been discovered, and the resulting improvement of the game had been recognized. The excellence of the game because it depended upon the intellect alone is already praised in the Middle Persian Chatrang-nāmak.

With the adoption of a rule of procedure by alternate turns of a single move each, a rule that does not always obtain in Indian dice-games, the game was complete so far as concerns essentials, and players had a workable game of war. Whether its invention may be ascribed to the Buddhist disapproval of bloodshed, which suggested to some enthusiast the possibility of replacing actual warfare by a game, it is impossible to say. It is at least suggestive that we shall find the game first mentioned in India in connexion with a stronghold of Buddhism, and that other early references will be associated with Buddhist regions.

The date when it occurred to some Indian to represent the chaturanga and its evolutions in a game cannot be fixed, though naturally it cannot be earlier than the organization of the army on which it is based. Chess was certainly in existence in the 7th century A.D., and it had already at that time penetrated to Persia. The evidence upon which the same has been asserted of China is unsatisfactory. The silence of Greek writers as to its existence, although after the time of Alexander the Greeks enjoyed an uninterrupted intercourse with India for two centuries, has been claimed by v. d. Linde (i. 78) as evidence for the non-existence of both the game of chess and also the ashṭāpada at that time, and although his conclusion has been disproved as far as the ashṭāpada is concerned, it is probably correct as regards chess. Writers who romance of ‘five thousand years ago’ and the like are indulging in mere speculation; the real position has been well put by Prof. D. W. Fiske:

‘Before the seventh century of our era, the existence of chess in any land is not demonstrable by a single shred of contemporary or trustworthy documentary evidence….. Down to that date it is all impenetrable darkness.’⁴⁷

The foundations of the modern investigations of early Indian literature for references to chess were laid by Prof. Albrecht Weber (B. 1821, D. 1901) in a series of papers read before the Berlin Royal Academy of Science in 1872–4. Before his attention was directed to the question by v. d. Linde, the only Sanskrit passage known to relate to chess was one which was first given in translation by Sir William Jones (B. 1746, D. 1794) in his essay On the Indian Game of Chess (Asiatic Researches, London, 1790, ii. 159–65). This gave a description of a four-handed dice-chess, and according to his informant, the Brahman Rādhakant, the Sanskrit text was an extract from the Bhavishya Purāṇa. Sir William Jones himself regarded this game as a modification of the primitive two-handed non-dice chess.⁴⁸ The exaggerated views current in the early part of the 19th century with regard to the antiquity of Sanskrit literature necessarily led to similar views regarding the age of this four-handed game, and Captain Hiram Cox propounded a new view in his paper On the Burmha Game of Chess (Asiatic Researches, London, 1801, vii. 486–511) by claiming that this four-handed game was the rudimental game of chess, and that the two-handed game was a modification of it. In the hands of Prof. Duncan Forbes (B. 1798, D. 1868)⁴⁹ this opinion was further developed into a complete theory of the development of chess. Briefly stated, the Cox-Forbes theory is this: A primitive four-handed dice-chess was practised in India about 5,000 years ago. As a result of the action of certain rules, or from the difficulty of always securing a full quota of players, the game gradually became a two-handed game. At a later time the civil and religious ordinances against the use of dice led to the abandonment of the dice-character of the game; and finally, by a rearrangement of the pieces, the game of chess as known to the Persians and Muslims came into existence.

In its inception this theory depended solely upon the supposed priority of the evidence for the existence of the four-handed game, and when Weber showed the unsatisfactory nature of the evidence in support of the statement that the Indian text was derived from a Purāṇa, scholars abandoned the theory altogether. In any case the 5,000 years of Forbes would have to be reduced greatly in view of the fact that modern scholarship does not place the Purāṇa earlier than 500–550 B.C.

We possess three texts of the passage in question,⁵⁰ which, however, all appear to go back to the same source, the Tithyāditattvam (Tithitattva) of Raghunandana, a writer of the late 15th or early 16th century. All are written in the Bengali dialect of Sanskrit in which the remainder of this legal work is composed. Weber claimed that there was nothing to show that the account is not an integral part of Raghunandana’s own book. On the other hand, as will be evident from an examination of the translation which I give in Chapter III, the text of the passage is defective towards the close, and the verses appear to be disarranged. This looks as if Raghunandana had used an earlier source, though since the three existing texts all show the same lacuna and preserve the same order, we are probably right in regarding the Tithitattva as the immediate source of our knowledge of the passage. For the view that the ultimate source is a Purāṇa, we have only the bare word of the Brahman Rādhakant.

When Weber wrote his papers, the Bhavishya Purāṇa was not accessible to European scholars. Several MSS. are now known to exist in India, and the work has been printed at Bombay (2 vols., 1897), but this edition is of no value for purposes of exact scholarship, as the editors have made extensive additions on their own responsibility. More useful are two MSS. now in the Bodleian Library, Oxford, of which Aufrecht has given good analyses in his Catalogue of the Sanskrit MSS. in that library. He makes no mention of any chess passage, and there is no connexion in which it might conceivably occur. Weber had already stated that later works based upon the Bhavishya Purāṇa, as the Bhavishyottara Purāṇa, contain no chess passage. And the silence of all other Sanskrit works before 600 A.D. makes Rādhakant’s assertion improbable in the highest degree.

Another theory of the ancestry of chess has been put forward by Mr. Culin in his Chess and Playing Cards (Washington, 1898). He sees in our present games the survivals of magical processes adopted in order to classify according to the four directions objects and events which did not of themselves reveal their proper classification. Dice or some similar agent represent one of the implements of magic employed for the purpose. According to his theory, chess is a game derived from a game of the race type, and the steps of the ascent are (1) two-handed chess; (2) four-handed dice chess (chaturājī); (3) Pachīsī, a four-handed race-game; (4) a two-handed race-game. It is therefore a development of the Cox-Forbes theory, which aims at carrying the pedigree still farther back. Culin’s argument is thus stated (op. cit., 858):

The relation of the game of Chaturanga (i.e. the four-handed dice-chess) to the game of Pachisi is very evident. The board is the square of the arm of the Pachisi cross, and even the castles of the latter appear to be perpetuated in the camps, similarly marked with diagonals on the Chinese chessboard. The arrangement of the men at the corners of the board survives in the Burmese game of chess. The four-sided die is similar to that used in Chausar (i.e. Chaupur). The pieces or men are of the same colours as in Pachisi, and consist of the four sets of men or pawns of the Pachisi game, with the addition of the four distinctive chess pieces, the origin and significance of which remain to be accounted for. By analogy, it may be assumed that the board, if not indeed all boards upon which games are played, stands for the world and its four quarters (or the year and its four seasons), and that the game was itself divinatory.

After stating that students of the history of chess do not now generally accept the Cox-Forbes theory, Mr. Culin continues:

Apart from this discussion, the relation of chess to an earlier dice-game, such as Pachisi, appears to be evident. The comparative study of games leads to the belief that practically all games as Chess, played upon boards, were preceded by games in which the pieces were animated by dice, cowries or knuckle-bones, or by staves, as in the Korean Nyout, the Egyptian Tab, and many aboriginal American games.

All students of the history of games owe very much to Mr. Culin for his careful investigations into the nature, implements, and rules of existing games. His suggestion that race-games may have originated in magical processes deserves consideration,⁵¹ and there is much to be said for his view that dice-games preceded games of pure combination. But neither hypothesis has as yet been established as fact, and the further step in his argument which deals with the connexion of the war-game chess and the race-game pachīsī is a very weak one. It has yet to be established that pachīsī or chaupur is older than chess.⁵² Mr. Culin’s argument depends too much upon resemblances which are only superficial, or can be explained equally satisfactorily in other ways. It shows signs of insufficient acquaintance with the known facts of chess history.

The theory that chess is a development of an earlier race-game involves the hypothesis that some reformer changed the whole nomenclature in order to make it self-consistent as a war-game, and secured the agreement of all his contemporaries. I find this hypothesis incredible.

Table of Contents

The earliest references in Subandhu, Bāṇa, &c.—The chess-tours in Rudraṭa.—Position in India c. 1000.—Some Arabic references.—Later Indian references.—Nīlakaṇṭ·ha.

Allusions to chess begin to appear in Sanskrit literature with the seventh century of our era, and a number of passages from works of that period have been discovered which have been held by Sanskrit scholars to contain references to chess. They vary considerably in value, and only one or two are sufficiently definite to convey any information as to the character of the game mentioned. In others, the only foundation for the belief that chess is intended is the use of the term ashṭāpada. Since this may equally well mean the older dice-game on the ashṭāpada board, these allusions cannot be conclusively attributed to the younger game of chess.¹

The earliest of these references occurs in Subandhu’s Vāsavadattā (ed. Hall, 284), a prose romance, written according to Macdonell (Skr. Lit., 232) ‘quite at the beginning of the seventh century’, which tells the popular story of Vāsavadattā, the Princess of Ujjayinī, and Udayana, King of Vatsa. In this work Subandhu thus describes the rainy season:

The time of the rains played its game with frogs for chessmen (nayadyūtair), which, yellow and green in colour, as if mottled with lac, leapt up on the black field (or garden-bed) squares (koshṭhikā).

The reference to chess in this passage appears to me to be quite satisfactory, although neither the name of the game nor the chessboard is mentioned. Had the race-game been intended, the men would almost certainly have been called sāri: the term nayadyūtair, which Thomas translates chessmen, is explained by the commentator as referring to chaturanga, and the comparison of the frogs hopping from plot to plot to the lac-stained chessmen moving from square to square is not inappropriate. From the mention of two colours only we may perhaps infer that Subandhu was thinking of a two-handed form of chess. Quite as interesting is the use of the word koshṭ·hikā, a cognate of koshṭ·hāgāra, for square. This word, meaning literally store-house or granary, is generally used in the sense of house, and thus presents a complete parallel to the Arabic bait, house, and the Italian casa (French case), house, which are both used in chess in the technical sense of square of the board. It has sometimes been suggested that the Sanskrit term was used as a result of the well-known Arabic legend of the reward bestowed upon the inventor of chess, a calculation which is so thoroughly Indian in character that it may be supposed to be much older that the earliest record of it now existing. It is more likely, I think, that the name koshṭ·hikā suggested the calculation of the sum of the grains of wheat than that the calculation suggested the name for the square of the board.

F. W. Thomas was the first to call attention to this passage in the ZDMG. (lii. 271). In a later note (ibid., liii. 364) he called attention to the use of the word varshākāla, ‘time of the rains’, or ‘the rains as Kāla’, and endeavoured to establish the reference to Kāla as a technicality of the game. As his argument is based upon the assumption that the Indian chessboard was already chequered in Subandhu’s time, it loses any weight it might otherwise have had. The chessboard has only begun to be chequered in Asia in our own time as the result of European influences. If the reference to Kāla has anything behind it, it is probably nothing more than the old and widely spread commonplace that fate plays its game with men for pieces.²

Slightly later than Subandhu is Bāṇa, who lived in the early part of the seventh century. Several possible references to chess have been discovered in his works by Macdonell and Thomas. Macdonell first called attention in the Athenaeum (July 24, 1897) to a passage in the Harshachārita, ‘the earliest attempt at historical romance in Indian literature’, in which Bāṇa gives an account of Srīharsha (Harshavardhana), the famous King of Kānyakubja,³ and supreme ruler of Northern India from 606 to 648 A.D., under whose patronage the work was produced. The passage contains a number of puns, and among others Bāṇa in describing the peace and good order of the realm remarks (Bombay edn., p. 86, 1. 11; Kashmir edn., p. 182, 1. 1) that

under this monarch (Srīharsha) … only bees (shatpada) quarrel in collecting dews (dues); the only feet cut off are those in metre: only ashṭāpadas teach the positions of the chaturanga.⁴

This reference seems to me particularly clear, and the rhetorical figure (parisankhyā) employed is admirably illustrated by the play on the two meanings of the word chaturanga. The mention of the name of the game, chaturanga, makes it plain that in this passage the word ashṭāpada is used in its original sense of a game-board, and not as the name of a game.

Thomas (ZDMG., lii. 272) has pointed to another passage of a highly figurative character in the same work. In this Bāṇa (Bombay edn., p. 10, 11. 10–12; Kashmir edn., p. 20, 11. 5–8; Eng. trans., p. 6) describes an angry sage as

contracting a frown which, as if the presence of Kāla had been obtained, darkened the ashṭāpada of his forehead, and was the crocodile ornament which bedecks the wives of Yama.

The scholiast explains ashṭāpada as chaturangaphalaka, i.e. the chessboard, but there is nothing in the passage itself to require chess. The simile would be suggested by the resemblance between the deep furrows on the brow of the angry sage, and the dividing lines of the game-board. Thomas suggested an explanation depending on the ‘mottled squares of the chessboard’: this is of course an anachronism.

Two passages also from Bāṇa’s Kādambarī have been cited as possibly containing references to chess. In Redding’s English version they are thus translated:

dice and chessmen (sāryaksheshu) alone left empty squares (p. 6),

and

Chandrapida went away at her departure followed by maidens sent for his amusement by the poetess at Kādambarī’s bidding, players on lute and pipe, singers, skilful dice and draught (ashṭāpada) players, practised painters and reciters of graceful verses (p. 152).

I do not think that we can accept either of these allusions as relating to chess. The use of the word sāri in the earlier passage makes it practically certain that a race-game of the pachīsī type is intended. In the second there is nothing to exclude the possibility that the older ashṭāpada game was intended.

Much more certain are the two references from Kashmirian poets of the ninth century which Jacobi gave in the ZDMG. in 1896 (1. 227 ff.). The earlier of these occurs in the Haravijaya or Victory of Siva (xii. 9), an extensive mahākāvya or artificial epic, by Ratnākara, a poet who mentions Bālabrihaspati or Chippata-Jayāpīda, King of Kashmīr, 837–47, as his patron, and whom a later writer, Kalhaṇa (Rājataranginī, v. 34), states to have been celebrated under Avahtivarman, 857–84. The chess passage is worded with the double meaning that was so favourite a device of the later Sanskrit poets. The poet is speaking of Aṭṭahāsa, one of Siva’s attendants, and if we read the passage one way it describes him as one

who continually turned the enemy in spite of the latter’s four-square force, of his abundance of foot-soldiers, horses, chariots, and elephants, and of his skilled operations with peace (sandhi) and war (vigraha), into one whom defeat never left (anashṭa-āpadam).

When read another way it may be translated—

who turned not into a chessboard (an-ashṭāpadam) the enemy who had a four-square (chaturasra) form, who abounded in foot-soldiers (patli), horses (ashwa), chariots (rat·ha), and elephants (dvipa), and who had the form (vigraha) of combination (sandhi),

i.e. according to Jacobi (op. cit., 228) and Macdonell (JRAS., 123), of two halves folding together, with reference to the symmetry of the arrangement. There can be no doubt from the mention of the four members along with the ashṭāpada that chess is intended, notwithstanding the non-use of the word chaturanga. The commentator, Alaka, son of Rājanaka Jayānaka, who probably lived in the 12th c., so understood it, for he explains ashṭāpada as chaturaṅgaphalaka.

The second passage is from the Kāvyālaṅkāra, a work by a slightly later writer, Rudraṭa, who is ascribed to the reign of Sankaravarman, 884–903 (adhyaya 6). He is enumerating different kinds of stanzas, composed to imitate the forms of various objects, and speaks (v. 2) of verses which have the shapes of

wheel, sword, club, bow, spear, trident, and plough, which are to be read according to the chessboard squares (chaturangapīṭ·ha) of the chariot (rat·ha), horse (turaga), elephant (gaja), &c.

The commentator Nami, who dates his work 1125 Vikr. = 1069 A.D., and who lived in Guzerat, explains chaturangapiṭha as chaturangaphalaka, and adds the comment ‘known to players’, and etc. as nara, by which we are to understand the foot-soldier (patti).⁵

1. Knight’s Tour (Rudraṭa).

2. Rook’s Tour (Rudraṭa).

3. Elephant’s Tour (Rudraṭa).

Rudraṭa next goes on to give examples of these metrical puzzles, and Jacobi discusses the three chess-puzzles at considerable length. The principle of construction is as follows: certain syllables are placed in the various squares of a half chessboard in such a way that whether the syllables be read straight on as if there were no chessboard, or be read in accordance with the moves of a particular piece the same verse is obtained. The ability to frame such puzzles argues considerable acquaintance with the moves of the chess-pieces, and the metrical conditions of the puzzle add largely to the difficulty of construction.

There is no difficulty in the cases of the rathapadapātha (chariot or rook tour) and the turagapadapātha (Knight’s tour). With the help of the commentator the solutions are easily ascertained. The move of the Turaga or Horse is identical with the existing move of the Knight. The Rat·ha’s move also is consistent with the existing move of the Rook. Both tours are so constructed that they can easily be extended to cover the whole board. Jacobi (op. cit., 229) notes that the Knight’s tour appears to have been very popular, since the commentator Nami gives a sloka which names the squares of the chessboard by akshara ha to sa.

The gajapadapātha, or Elephant’s tour, presents considerable difficulty. In the first place a complete tour is impossible of construction with the move ordinarily associated with the Elephant (Bishop) in early chess. We have accordingly to do here with an unusual move. If we examine the commentator’s solution, exhibited in diagram 3 above, we see that it consists of two halves, each occupying two lines of the board, that the two halves are precisely the same, and that they are connected by a move from h7 to a6, right across the board. Jacobi treated the diagram as containing two separate solutions, each being an Elephant’s tour upon two lines of the board, and ignored the abnormal leap that apparently connects them as inconsistent with any move ever used in any ordinary game of chess. He then shows that the moves in these two tours are consistent with a fivefold move which al-Bērūnī records as in use in the Punjab in his time, which is still the Elephant’s move in Burmese and Siamese chess, and which occurs in Japanese chess as the move of the differently named piece which occupies the same initial position as the Elephant in most varieties of chess. This move was one to the four diagonally adjacent squares and to the square immediately in front; see diagram 3 on p. 59. Jacobi’s explanation is, however, met by the obvious objection that such a move can easily be extended to cover the half board without the necessity to use an abnormal leap, and it is necessary to explain why it happened that Rudraṭa did not complete his tour in an orderly way when apparently possible, before we can accept the explanation. The fivefold move only admits one possible chess solution which is distinct from a Rook’s tour, viz. that of the diagram on this page, where the lower rows repeat the tour of the upper rows in the reverse direction. Rudraṭa’s problem, however, is not solely or even in the first case a chess one, but is governed by difficult metrical conditions—the syllables must give the same reading whether read as written or read in accordance with the chess rules. A brief examination of the diagram on this page shows that the tour there described allows the use of only two different syllables in the third and fourth lines; thus aababba, abbbabaa. The composer has to replace a and b by two syllables which will afford an approach to a meaning when arranged according to this sequence. Such a task approaches sufficiently near to impossibility to justify the abandonment of the chess condition in part; the composer has carried out a task of quite sufficient difficulty in providing two different metrical solutions for the tour over the two lines.⁶

A still later allusion to chess occurs, as Weber pointed out,⁷ in Halāyudha’s commentary on Pingala’s Chandaḥsutra, which belongs to the end of the tenth century. Halāyudha is discussing the form of certain metres, and incidentally instructs the reader to

draw a table of 64 squares (koshṭhāgara) as in the game of chaturanga.

These passages include all the known references to chess in Indian literature prior to the year 1000. We cannot claim that they establish much beyond the existence of the game, or that we have travelled far from the ‘impenetrable darkness’ of the earlier period. We can, perhaps, form some opinion of the spread and popularity of the game in India from these allusions. We find chess specially connected with the North-West of India, and the upper basin of the Ganges; we find it sufficiently well known in the 7th c. in this region for it to furnish comparisons to the poets and romancers of the time, and so well known in Kashmīr in the 9th c. that not only did poets employ similes derived from its special features, but that the ingenious also devised complicated and difficult puzzles which depended for their solution upon a practical knowledge of chess. The commentator on these puzzles shows that in the 11th c. the game was known in Guzerat, so that by that time we can safely assert that a knowledge of the game was common to all Northern India. The same century may have seen chess practised in the Deccan, if Dr. Bühler’s statement that the Mānasollāsa of the Ṣālukya (Solanki) Prince Someṣvara mentions chess among his recreations can be proved to be accurately translated.⁸ It is not clear whether chess had reached the South of the peninsula in the year 900, for the Arabic traveller, Abū Zaid as-Sīrāfī,⁹ when describing the gambling habits of the inhabitants of the coast opposite Ceylon, only alludes to nard and cock-fighting among their recreations. If, however, the date assigned to the Sinhalese commentator to the Brahma-jāla Sutta is correct, chess cannot have been much later in reaching the South of India and Ceylon.

The oldest foreign references to the practice of chess in India occur in Arabic works. Two of these are of great importance, for in place of the usual Arabic legends of the invention of chess which will be discussed in a later chapter, they give us more or less detailed accounts of the game as it was played in India at the time these works were compiled.

The earlier of these is a short note which probably formed part of the lost chess work of the Arabic master al-‘Adlī, who was at the height of his fame about 840 A.D. The note is preserved in two later MSS. based in part upon al-‘Adlī’s work, of which I have made great use in my chapters on the Muslim chess. In AH (f 24 a = C f 33a) the note concludes the section on derivative games which is introduced by the rubric ‘Al-‘Adlī has said’, which throughout the MS. precedes extracts from this writer. In H (f 20)¹⁰ the note is given in a much condensed form, but again concludes the same section from al-‘Adlī’s book. The passage in AH runs as follows:

And this form is the form of chess which the Persians took from the Indians, and which we took from the Persians. The Persians altered some of the rules, as is agreed. It is universally acknowledged that three things were produced from India, in which no other country anticipated it, and the like of which existed nowhere else: the book Kalīla wa Dimna, the nine cyphers with which one can count to infinity, and chess. The Indian claim to Astrology and Medicine is disputed by the Persians and Greeks.

Of the Indian rules of chess, one is observed by the people of Ḥijāz, and is called by them the Medinese Victory. If there be with the Kings two pieces, and the King can take a piece, then which ever first takes, so that the other is left with nothing, wins: for the other side will have been left at a particular time destitute of comrades. This is an Indian rule according to which the people of Medina play.

Another Indian rule is that when the King cannot find a square into which to move, and the other King has nothing wherewith to checkmate him, the first has won. But this is not a Persian rule.¹¹

Another Indian rule is that the Elephant is placed in the corner, and omits one square in a straight line to jump into the second in a straight line. And this it does in all the squares of the board. Each Elephant has 16 squares, and the company of Elephants can get into all the squares without collision. But in the form of chess which we have taken from the Persians, and which is played now, the Elephants have only half the board, and each Elephant has 8 squares. The number of squares has been reduced because they go slantwise.

An Indian was asked why they put the Elephant in the corner, and replied that the Commander of an army in which there are elephants must, owing to his importance, be given the place of commander of either the right or left wing. The Persians, however, think that he should be put next the King, being required for pursuit or flight. The Rooks, he said, are horses in … (a lacuna, after which the writer goes on to praise the horse and falcon, and discusses the relative precedence of the kings of Babylon, India, China)…. The value of the Indian Elephant is the same as that of the Firzān (counsellor, the mediaeval Queen).

The second account is to be found in al-Bērūnī’s India. The author, Abū’r-Raiḥān Muḥammad b. Aḥmad al-Bērūnī, was born at Khiva in Khwārizm in 362/973 and lived in Hyrcania on the Southern shores of the Caspian. He died at Ghazna 440/1048. He travelled into India but penetrated no farther than the Punjab, and, besides other works of a historical and chronological character, he wrote c.421/1030 an account of the religion, philosophy, literature, chronology, astronomy, customary laws, and astrology of India. His work is an extremely valuable record by a keen inquirer, but unfortunately he appears to have brought away a rather hazy impression of that variety of chess which was peculiar to India. In this, however, he is no worse than the vast majority of observers even in modern times. He says:¹²

In playing chess they move the Elephant straight on, not to the other sides, one square at a time like the Pawn, and also to the four corners like the Firzān. They say that these five squares—i.e. the one straight forward, and the others at the corners—are the places occupied by the trunk and the four feet of the Elephant.

They play chess, four persons at a time, with a pair of dice. Their arrangement of the figures on the chessboard is the following:

As this kind of chess is not known to us, I shall explain what I know of it. The four persons playing together sit so as to form a square round a chessboard, and throw the two dice in rotation. Of the numbers of the dice the 5 and 6 are not required. Accordingly, if the dice show 5 or 6, the player takes 1 instead of 5, and 4 instead of 6, because the figures of these two numerals are drawn in the following manner—

so as to exhibit a certain likeness of form to the 4 and the 1 in the Indian cyphers.

The name of King applies here to the Firzān (Minister).

Each number of the dice causes a move of one of the figures. The One moves either the Pawn or the King. Their moves are the same as in the common chess. The King may be taken, but is not required to leave his place.

The Two moves the Rook. It moves to the third square in the diagonal direction, as the Elephant moves in our chess.

The Three moves the Horse. Its move is the generally known one to the third square in the oblique direction.

The Four moves the Elephant. It moves in a straight line, as the Rook does in our chess, unless it be prevented from moving on. If this be the case, as sometimes happens, one of the dice removes the obstacle, and enables it to move on. Its smallest move is one square, its greatest 15 squares, because the dice sometimes show two fours, or two sixes, or a four and a six. In consequence of one of those numbers, the Elephant moves along the whole side on the margin of the chessboard: in consequence of the other number it moves along the other side on the margin of the chessboard, in case there be no impediment in the way. In consequence of these two numbers the Elephant in the course of his move occupies the two ends of the diagonal.

Four-handed chess. After al-Bērūnī.

The pieces have certain values, according to which the player gets his share of the stakes; for the pieces are taken and pass into the hands of the player. The value of the King is 5, that of the Elephant 4, of the Horse 3, of the Rook 2, and of the Pawn 1. He who takes a King gets 5, for two Kings he gets 10, for three Kings 15, if the winner is no longer in possession of his own King. But if he has still his own King, and takes all three Kings, he gets 54—a number which represents a progression based on general consent, and not on an algebraic principle.

In the main this is a description of the four-handed dice-chess to which I devote the next chapter. Falkener (139–42) thought that al-Bērūnī only refers to this game, and that he never saw the two-handed game in India. But Falkener treats al-Bērūnī in a very cavalier manner, going so far as to declare that he can have been no chess-player. On the other hand Sachau, Gildemeister, v.d. Linde, and v.d. Lasa all agree in thinking that al-Bērūnī did see both games in India, and the last two writers think that it is possible to infer from his describing the four-handed game in terms of the ordinary chess, that he regarded the former game as a modification of the latter. This seems to be going too far: al-Bērūnī, writing for Arabic readers, would naturally explain the Indian game by comparing it with the Muslim game that his readers knew. But I think it is quite clear that al-Bērūnī did see the two-handed game in India, firstly from the fact that he gives two descriptions of the Elephant’s move; secondly from the curious clause that the name of the King applies also to the Firzān. Four-handed chess is still played in India, and it is usual to use the ordinary set of chessman for the purpose. The two allies share out the men of one colour, and one uses the ‘Queen’ as a King. I believe that the clause refers to this custom, and that it accordingly presumes the existence of ordinary chessmen and consequently a knowledge of the two-handed game.

1. Indian Four-handed chess.

2. Indian (al-‘Adlī).

3. Indian (al-Bērūnī).

The Elephant’s Move in early Indian chess.¹³

The fivefold move of the Elephant has been felt to be a difficulty. Falkener suggested that al-Bērūnī must have obtained it from Japanese chess! But there was no necessity to go so far afield. The move exists in the Burmese and Siamese games, and Rudraṭa’s tour raises the presumption that it existed in the Punjab or at least in Kashmīr before al-Bērūnī’s visit. Moreover, the al-‘Adlī account shows that the move of the Elephant was not fixed in India. We have records of no less than three moves of this piece having been tried in India, and with the discovery of this uncertainty the difficulty that has been felt ought to disappear.

These three moves are exhibited in the diagrams on this page. The first, a diagonal leap, became the widest spread, and it is probable that it is the oldest move. It is the only one which passed westwards, and it exists in Chinese chess also. It became again at a later date the ordinary Indian move. Al-Bērūnī records it as existing in the four-handed game, though in connexion with the Rook. The appearance of the other two moves may have been due to a feeling that the original move was not in harmony with the value of the elephants in war. In actual life they were highly esteemed as one of the most potent divisions of the army; on the chessboard it must have soon become evident that the Elephant was the weakest of the major pieces. The obvious remedy for this want of verisimilitude was to increase the power of move of the chess-piece. Al-‘Adlī records one such attempt. The power is evidently increased, twice as many squares are now accessible to each Elephant, and one or other of the four Elephants on the board can now reach each of the 64 squares; the power is now estimated to be equal to that of the Firzān (counsellor). The attempt which al-Bērūnī records appears to be a later one, and it has proved more enduring. It has the advantage of fitting in with the peculiarly Indian idea that the elephant is a five-limbed animal, which has resulted commonly in the description of the trunk as a hand. The move also gives the piece a higher value which has been estimated as rather more than that of a Knight. This move appears to have been in the main associated with Buddhist centres, and its disappearance from India may be connected with the overthrow of Buddhism there.

Al-‘Adlī’s statement that in India the Elephants occupied the corner squares is the earliest reference to the uncertainty in the position of this piece, to which I have already referred. From a comparison of the existing information the following points become clear.

(1) In the four-handed game the piece with the Rook’s move stood next the King, and the piece with the Elephant’s move stood in the corner. The piece next the King retained the name of Elephant.

(2) Two authorities (al-‘Adlī and the late Vaidyanātha, see later) transfer this arrangement of the moves to the ordinary chess, so that the piece with the Rook’s move stood next the King, and the piece with the Elephant’s move stood in the corner. In these cases the names were also interchanged, and the Elephant stood on a1, &c.

(3) By the 17th c. generally the piece with the Rook’s move had been definitely fixed on the corner squares, but changes were introduced in the nomenclature. To-day three main divisions may be made. The original nomenclature, Chariot a1, Horse b1, Elephant c1, is the usual nomenclature in Northern India and in the Maldive Islands. The inverted nomenclature, Elephant a1, Horse b1, Chariot e1, is the rule in the extreme South of India among the Tamils, Telugus, and Kannadis. A new nomenclature, Elephant a1, Horse b1, Camel c1, is widely spread. It has been noted as far North as Delhi, and is the rule over the greater part of Central India and the Deccan.

From al-‘Adlī we learn that the Indian rules varied in two particulars from those of Baghdad. One of these variations relates to Stalemate, a situation without parallel in war, which is a consequence of the limited area of the board, and the method of play by alternate moves. The rules regarding Stalemate have varied all through the history of the game, and this old Indian rule by which the victory is given to the player whose King is stalemated, illogical as it is, reappeared in England from 1600 to about 1800. In India the rule has long been replaced by other conventions.¹⁴

The other relates to the ending which, following the usage of early English chess, I call Bare King. In early chess the player who was robbed of all his men lost the game. Occasionally it happened that at the close of a game both sides were reduced to the King and a single piece, while the player whose turn it was to move could take the enemy’s last piece, leaving his own piece en prise. Indian—and Medinese—players counted this a win to the first player on the ground that the opponent was first bared. Persian, and Arabic players generally, reckoned such an ending as drawn.

Chess must have received a great stimulus in India as a result both of the Muhammadan invasion and conquest of North-Western India, begun before 750 and completed by 1100, and of the settlement in South-West India of Persian (Parsi) refugees in search of an asylum where they could still practise their Zoroastrian religion. But while the Parsis appear to have adopted the native Indian method of play, the Muslim conquerors brought with them their own game, and have retained it ever since almost entirely free from Indian influence. It is probably due to this Muslim conquest that the references to the ordinary two-handed chess that I have been able to collect for the 11th to 18th centuries are drawn entirely from Central and Southern India.

It is a very remarkable fact that in these Southern works, chess, the two-handed game of pure combination, is no longer called chaturanga, but has received a new name. The exact form of this name varies from one authority to another, but in every case the word is a compound of the Skr. buddhi, intellect, and all the forms may be translated by the one English name, the Intellectual Game.¹⁵ But it is perhaps even more remarkable that the name chaturanga appears side by side with the new name of chess as the name of a dice-game. It has generally been assumed that this was a two-handed dice-chess, but this does not seem to have been the case. All the evidence goes to show that this dice-chaturanga was a game closely allied to the original ashṭāpada game, if not that game itself.

I imagine that the explanation of this strange transference of name is as follows. The invention of chess did not interfere with the popularity of the ashṭāpada game, and for a long time the games existed side by side, the race-game preserving its old name, and chess being known as chaturanga. Gradually the term ‘ashṭāpada’ passed out of use: we have already seen how commentators of the older literature found it necessary to explain ashṭāpada by chaturangaphalaka, chessboard. At the same time the original meaning of ‘chaturanga’ was forgotten and the word was known in colloquial language merely as the name of a game, the game played on the chaturangaphalaka. The time then came when—possibly only in Southern India, far from the original home of chess—‘chaturanga’ was used indifferently for both games played on the chessboard. With the necessity for discrimination between two games so different in character, the name ‘chaturanga’ became confined to the more popular game, which happened to be the race-game, and a new name had to be found for the less popular game, chess. A name was chosen which admirably described the distinctive feature of chess, its freedom from the sway of chance, and its presentation of a struggle between two minds for the mastery. To-day chess is practically unknown to the natives of Ceylon, but the race-game on the board of 9 × 9 squares is known in Ceylon and Southern India as Saturankam or Chaturanga.¹⁶

This Southern Indian use of chaturanga as the name of a race-game provides a satisfactory explanation of certain statements by commentators which have hitherto puzzled chess-writers. Thus Govardhana (12th c.) in his Saptasatī mentions a poor woman who lives and dies, tormented by the fire of separation, and revives again at a kind look from the eye of the villain (lit. player, but the word had obtained the derived sense of villain from the unfair play that the gambler so often employed) like a sāri. The commentator Ananta (1702 without era, therefore either 1646 or 1780) adds, ‘i.e. like a chaturanga-man (chaturaṅgagūṭikā, lit. chess-horse), which, as often as it dies, i.e. is placed out of the game, is always again restored by the fall of the dice.’ Similarly, the undated commentator to Dhanapāla’s Rishabhapaṅchāsikā (c. 970 A.D.) explains the obscure passage—‘The living beings become like sāri on the board (phalaka) of life, although torn from the senses (i.e. set in motion by the dice) if they espy you (the point of the board) not sharing in imprisonment, murder and death’—as referring to chaturanga. For Dr. Klapp’s consequent mistake, see ZDMG., xxxiii. 465, and Qst., 5. The chaturanga of both these scholiasts is, I feel certain, the race-game, not chess.¹⁷

The same game is obviously intended in the passage quoted by Weber¹⁸ from a MS. of 1475 Samvat (= 1419 A.D.) of the Siṅhāsanaxatriṅṣika, in which a gambler discourses at length to King Vikramāditya on the different games that he knows and their special excellencies, among them being chaturaṃga.

Chess and this race-game chaturanga appear in sharp contrast in the Pañchadandachattraprabandha, a Jaina version of the tales of King Vikramāditya,¹⁹ which contains many Persian words and is not older than the 15th c. In the story the King is set the task of defeating the daughter of a wise woman thrice at play. The King offers her the choice of games, and like Yvorin’s daughter in Huon of Bordeaux, she prefers not to risk her reputation upon the chances of the dice.

The king said: ‘What game will you play?’ She answered, ‘What are the other games worth, rāmdhika, nāla, chashi, lahalyā, chaturaṃga-ṣāri, paṣika, &c.? We will play the intellectual game (buddhidyuta).’ ‘As you wish’, said the king. The king ordered a board (phalaka) to be brought; the game was arranged on both sides: Prince (nṛipa), Counsellor (mantri), Elephants (hasty), Horse (aṣva), Infantry (padāty), and Forerunner (agresara). They began step by step to play the moves (?), The king decided naturally upon an involved game, and he began to play with the help of his invisible āgnika.²⁰

The list of the pieces leaves no doubt as to the identity of buddhidyuta with chess. All the original members of the chaturanga are here except the Chariot, whose place is taken by the Forerunner (agresara). Weber (op. cit.) and Gildemeister (Schaakwerld, 1875, 330) see in the use of this term one of the Persicisms so frequent in the work, and recall the occasional use of the Per. mubariz, champion, as an epithet of the Rook in the Shāhnāma. But there is no evidence that the Persians ever gave the piece any name except Rukh, and this explanation has nothing to recommend it. I think we must regard it as entirely Indian. There has always been a greater variety in the names of the pieces in Indian chess than in the game elsewhere.

We have a very important section on chess at the end of the fifth book (the Nitimayūkha) of Bhaṭṭa Nīlakaṇṭ·ha’s great encyclopaedia of ritual, law, and politics, the Bhagavantabhāskara. This work was written about 1600 or 1700 at the command of Bhagavantadeva, son of Jayasinha. The fifth book treats of monarchs, their anointing and consecrating, the whole course of the royal method of life, and the instruments by which the king governs. One of these is the army (bala), and in this connexion Nīlakaṇṭ·ha permits himself a digression in which he speaks of the game which depends not on mere material force but on mental powers.²¹

1. After the discussion of the foregoing subject, viz. the deportment of kings, which is most important for princes, Nīlakaṇṭ·ha, the son of Ṣaṃkara, describes the intellectual game (krīḍā buddhibalāṣritā).

2. We draw eastwards 9 lines and also northwards 9 similar lines upon a piece of cloth, or on a board or on the ground. Thus we obtain the board of 64 squares (catuḥshashṭipadā).

3. We mark the corner squares with geese-feet, also the two middle squares in the same lines, also in the centre we mark 4 squares, and we arrange the warring forces of the two armies on the board.

4. On the two centre squares of the last 8 squares stand the King (rājā) and Counsellor (mantrī), by them the Camels (ushṭra), then the two Horses (vāha), then the two Elephants (dantī). In the next row are placed the 8 Pawns (patti). The host on the other side is arranged similarly, and both are ready for battle.

5. The King moves straight and aslant to 8 squares; the Counsellor aslant only; the Camel (karabha) moves similarly but it passes over a square in the middle like a chain; the Horse (vājī) passes over a square different from the square lying in the straight line into 8 aslant squares. The Elephant (kuñjarā) moves straight out to all squares in its file. The Pawn goes straight forwards.

6. It takes always moving obliquely. When it arrives at the last square, it becomes a Counsellor when it is returned thence to the square it occupied previously. If it arrive at the end on a goose-foot it becomes a Counsellor at once, and not only after the return to the former square. Thus the rule is correctly taught according to the regulation.

6*. Dividing itself by non-repetition, and variety, the game is doubly desired. There is a division for the square, and what is placed upon it, and through this the first is doubly desired. (Text corrupt, and meaning doubtful.)

7. Hereupon the two Pawns (padati) which stand before the two Counsellors (sachivā), and along after them the two Counsellors themselves are to be moved two squares distant. Also another piece which goes one square distant is advanced at the same time by others.

8. A piece standing in the way does not hinder the Horse (haya) and Camel (ushṭra) from going and coming. The Horse and the rest hinder the Elephant (gaja) if they stand before it.

9. The two Pawns (patti) which are placed next the back corners of the Counsellor are firm, so also are the two which go in the chain behind the Camel.

10. This army placed in double array which accomplishes the slaughter of the enemy according to the usual arrangement is called durokhaṣa.

11. If the Elephant (dvipa) is placed in the centre opposite to the opposing King after the removal of his own, it is called kātīṣa.

12. No piece should be placed without protection, and it is desirable to protect by a weaker piece. It is not proper to protect another piece rather than the King. The slaying of the King is yet considered proper.

13. Imprisonment is counted as a defeat of the King. If the King is left entirely alone it is reckoned a half-victory, if he is checked 64 times in succession he is also held to be defeated.

14. When a King is imprisoned without standing in check, and no other of his pieces can move, he may slay the piece of the enemy in his vicinity which imprisons him.

15. If a piece remains over in the army of the imprisoned King, the player of it counts up the counter-marks (?); then he adds 2 for himself and doubles the sum. (Meaning not clear.)

16. When he has finished, he numbers the marks, if there are 64 against him, he loses. If he has as many he is equally defeated, if he has more the result is reversed.

Immediately following this text are three Knight’s tours, the solutions of which are concealed by syllables written on the chessboard, which, when read in the correct order of the tour, yield a connected text. These tours are not only re-entrant, but also to a certain extent symmetrical, and the verses are all based on the same tour, starting from different squares.²² The text begins:—

Draw a diagram of 64 squares, write the syllables siṃ na hi beginning in the S.W. (top right-hand) corner, and also in the N.E. (bottom left-hand) corner. Afterwards move the Horse by reading these syllables, ṣri siṃ, hana, &c.

The solution to the first diagram, ascribed to a king of Sinhaladvīpa (Ceylon), is—

There was a rich host of wise men under king Ṣrī Siṇhaṇa. They knew how to move the Horse into every square, a move at a time.

The second diagram is ascribed to Nīlakaṇṭ·ha’s father, Ṣamkara.

Ṣamkara moved the Horse from his square by 63 leaps in the incomparable palace of Prince Rāmeṣa surnamed Nārāyaṇa.

The third diagram is solved by a poem, which concludes:—

Thus again Nīlakaṇṭ·ha moved his Horse from here.

It is accordingly Nīlakaṇṭ·ha’s own.

Knight’s Tour (Nīlakaṇṭ·ha).

It has generally been assumed that Nīlakaṇṭ·ha describes a game that has been largely influenced by Persian usages. This view depends mainly upon Weber’s clever conjecture that the two technical terms durokhaṣa and kāilṣa were Sanskrit transliterations of Persian terms—du-roka-shāh (two Hooks-King, i.e. the game in which these pieces have their usual positions) and kāt-i-shāh (the migration of the King, i.e. the game of transposed King and Rook). This, however, is entirely a matter of nomenclature, and I can detect no other evidence of Persian influence. The method of play is unlike that of the Persian Shaṭranj, and the rules are throughout essentially Indian. We may account for the two Persian technicalities by ascribing their introduction to Parsi players.

Nīlakaṇṭ·ha’s account of chess is on the whole clear and intelligible; the few obscurities only concern minor points, such as the method of calculating the result in the case of stalemate. The instructions for describing the chessboard are very interesting; the scratching of the diagram on the ground is contemplated, and the marked squares are carefully defined. Apparently the arrangement of the chessmen is the normal one, and the two Kings are placed upon the same file (see § 11). The want of fixity in the names of the pieces is typically Indian. The name of each piece is constant, but four different names are used for the Elephant, three for the Horse, and two each for the Camel, Counsellor, and Pawn. I infer from this that Nīlakaṇṭ·ha was accustomed to play with carved pieces which reproduced the actual figures of men and animals. The two players (7) commence the game by each making a double move: Pd4 and Qd3, Pd5 and Qd6. Some players moved a third man on this move, apparently a second Pawn. The initial double step (9) is only allowed to the Pawns on the a, d, and h files; the other Pawns can move only one square at a time. Promotion (6) is connected with the marked squares; the Pawn ‘queens’ at once on the marked squares a8, d8, e8, h8; but elsewhere it has to make some further move—apparently to the square it had occupied the previous move, but the text is not sufficiently explicit. Checkmate and Perpetual check are wins, Bare King a half win. A King in a position of stalemate is allowed to remove the piece which confines him: the final result of this position apparently varies with circumstances.

Nīlakaṇṭ·ha’s rules are important as the earliest statement of the rules of the native chess of Southern India. In some points his rules approximate to rules observed in Malay chess; in others they show a remarkable similarity with the rules associated with the German village of Ströbeck. In common with existing forms of Indian chess (specially the form I call Parsi chess) are the restrictions on the double step of the Pawn, and the abnormal method of playing the first move. In contrast are the rules of Pawn promotion.

Slightly later than Nīlakaṇṭ·ha is a work by Vaidyanātha Pāyaguṇḍa, who lived in the first half of the 18th century or later. This work has for title Chaturangavinoda, The Game of Chess, but only the last chapter of 44 slokas treats of the game. The text of the unique MS.²³ is hopelessly corrupt, and Weber could only give a few extracts. It deals with the ordinary two-handed game without dice. Beyond this we only know—

The Chariots (ratha, syandana) occupy the corners, next to them are the Horses (turaja). then the Elephants (dvīpa, nāgendra, nāga), and in the centre are the King (rājā, nripa) and his Counsellor (mantri). The 8 Foot-soldiers (padāti) stand in front….

The Chariot leaps diagonally into the third field….

The Horse goes (?) to the corners of a square standing on 4 squares….

The Elephant goes in the 4 streets….

The Counsellor goes one or two or all squares diagonally….

The King goes to all the squares round about….

The Pawn goes one field forwards, and takes to both sides….

The special points about this description are: (1) the name chaturanga is still used for the ordinary two-handed chess, (2) the original names of the pieces remain, (3) the Chariot and Elephant have interchanged moves, precisely as al-Bērūnī describes in the case of the four-handed game, and (4) the Counsellor’s move is approximating to the modern move of the Queen: it is apparently identical with our Bishop’s move.

I have now come to modern days, when Europeans were again coming into direct contact with India. We possess no satisfactory accounts of Indian chess in the descriptions of the early voyages to the East. A few sets of native chess were brought home, and Hyde obtained some from Sheldon and describes them in his Mandragorias.²⁴ Forbes (162–3 and 249–51) quotes from two English volumes of memoirs of the close of the 18th cent. some references to games between Europeans and natives, but the information is too unscientific to be of much value.

Table of Contents

Introductory.—Spread in Malay lands.—Early references.—The chessboard.—Nomenclature.—Moves of the pieces.—Rules.—Illustrative games.—Malay chessmen.—Concluding observations.

Although chess is known and played in every Asiatic country to the east of India, the forms of the game that are played by the different peoples present at first sight as wide differences as are found anywhere in chess. On closer investigation, however, it is possible to discover certain common features in some types which enable us to classify these games in three groups, corresponding to the known ethnological families and religions of Eastern Asia. To one group, comprising the chess of Burma, Siam, and Annam, three countries linked by that form of Buddhism which is conveniently called Southern, I devote Chapter VI; to a second group, comprising the chess of China, Corea, and Japan, I devote Chapter VII; while in the present chapter I shall deal with the varieties of chess current among the Malays, which are united by the phenomenon of a nomenclature which has been drawn from many sources, and by a type of move which is closely akin to that of modern European chess.

To-day, chess is very widely played by the Malay races, and ranks as one of their most popular games.¹ On the mainland we possess records of its practice in the British Straits Settlements (Malacca), in the Protected States (Selangor), at Kelantan, and at Johore. We also possess good descriptions of the game as played in Sumatra, in Java, and in Borneo. Von Oefele, who has made a most patient and valuable study of the game as played in Sumatra by the Orang-Batak,² records that practically every male Batak has some knowledge of chess, while nearly every village meeting-hut has a chessboard carved upon its wooden floor. So violent are the passions aroused at times by the game, which is always played for a stake, that the headman of the village has occasionally had to forbid the practice of the game for a season.^2a

We know very little of the history of chess in these lands. The few Europeans who have made any study of the early history of the Archipelago speak of four successive waves of foreign culture and religion, all of which have in turn left a notable impress upon the customs of these islands. Somewhere about the 7th c. A.D. the Hindu religion established itself in Sumatra and Java, to be followed by Buddhism, and rather later—from the 13th to the 15th c.—by Muhammadanism. From the beginning of the 16th c. the coastal regions have been in continual contact with Europeans, first with the Portuguese, afterwards with the Dutch, and at a later date still with the English. Malay chess reflects all these invasions, since it shows unmistakable traces of Indian and of Arabic, and also of European influence.

The game is certainly older than any European influences, for on the arrival of the first Portuguese expedition off Malacca in 1509 its commander, Diego Lopez, was playing chess when a Javan from the mainland came on board. The native recognized the game at once, and had some conversation with Lopez on the forms of chessmen used by his countrymen.³

There is also a reference to the game in the Sějarah Malayu, native history dating from the early 17th century, ch. xviii, in connexion with a visit to Malacca by a certain Tan Bahra, of Pasei in Sumatra.⁴ The passage goes on to say—

Now this Tan Bahra was a very skilful chessplayer, and one that was unequalled at the game in that age, and he played at chess with the men of Malacca … and beat them all: but Tan Pakarma, son of the Bandahara Paduka Raja, was able to make some resistance … and if Tan Bahra threw away a pawn at the corner, then he was beaten by Tan Pakarma.

Broadly speaking, all forms of Malay chess are played in the same way, the differences only appearing in the refinements of the game. It will therefore be simpler to treat of the game as one, and to deal with the variations of rule or practice as they arise. Even in Java, where for some unexplained reason the otherwise universal Malay nomenclature is replaced by another, the game remains practically the same as in the other parts of the Malay world.

The game of chess has two names in Malay. The commoner name is main chātor,⁵ in which main is the Malay word for game, and chātor can hardly be anything but a broken-down form of the Skr. chaturanga. This name is the only one recorded for Borneo, Java, and the Batak race. It is given as the ordinary name by all my authorities except Dr. Marsden,⁶ who both in his History of Sumatra (ed. 1811, 273) and in his Malayan Dictionary (Part ii, Eng. and Malayan, s.v. chess) only gives the name as main gājah, i.e. the game of the elephant. This name has been recorded as used on the mainland both by Robinson and by v. Oefele, who gives it in the form permainan gājah. This form may be due to the influence of the name of another favourite game, the main rimau, or ‘game of the tiger’,⁷ though it is not easy to see why the Elephant should have been selected for mention, rather than the Horse or any other piece. The hypothesis that it may be due to Chinese influence—which is based upon the presence of Chinese settlers on the sea coast of the Peninsula, and all the islands, and upon the fact that one possible meaning of the Chinese name of chess, siang k‘i, is ‘the game of the Elephant’—must be rejected, because in all other cases of cultural borrowings, the Malays have adopted, and not translated, the Chinese name. Moreover, it is not easy to see why the Chinese chess, which does not appear to the casual observer to have any connexion with the Malay game, should have been able to exert an influence which was at once so strong that it led to the introduction of a new name for the game, and so weak that it left the actual method of play absolutely untouched.

Malay Chessboards. Skeat Collection, Museum of Archaeology and Ethnology, Cambridge.

Malay Chessboard.

(Malacca and mainland.)

Malay Chessboard.

(Sumatra.)

The Malay chessboard (lōh chātor or pāpan chātor⁸) is un chequered, but exhibits special markings which are characteristic of all Oriental boards. These vary so much in the few Malay boards which I have seen that it is clear that no traditional arrangement survives, and I think it probable that they have often become merely decorative. Some of the arrangements are very like those of neighbouring countries; thus one of the boards in the Skeat collection at Cambridge has a traditional Indian marking while the other resembles one of the Burmese markings. The ordinary board of the mainland is said by Mr. Robinson to have only the main diagonals marked—again a Burmese marking—and these diagonals are connected now with the rules of Pawn-promotion, and have probably suggested them. In Sumatra the board has a far more intricate appearance, since the complete network of diagonals of all the 64 small squares of the board is inserted. V. Oefele explains their presence as arising from the method of constructing the board. In order to obtain the correct proportions, he says, the Batak player first draws the outer square, then he inserts the diagonals to obtain the centre of the board and draws parallels to the sides through this point. By repeating this method he obtains accurately the quarter board and the eighth, and so obtains his 64 squares all of a size. Finally, to preserve the symmetry, he adds the missing diagonals, and the complicated figure is complete. This explanation does not seem to me to be satisfactory: while it certainly gives a convenient way of producing the final Sumatran figure, it is by no means the most natural way to draw a board of 8 by 8 squares.

The board is often made of wood, with the lines incised. This may be done upon a board of the floor of the hut, and a board for permanent use may be so secured. But it is also often scratched in the ground for an alfresco game when a movable board is not at hand.

The two nomenclatures may usefully be contrasted thus:

Of the ordinary Malay names, rāja (= king), mantri (= counsellor, minister), and gājah (= elephant) are all Sanskrit words, and we have already seen that they are or have been in regular chess use in India. Kūda (= horse), tēr (= chariot), and chemōr (= chariot) are Tamil and Telugu, languages spoken on the south-east coast of India, in the vicinity of Madras. The use of chemōr in chess in India has not yet been verified, but tēr is used as the name of the Rook in Tamil, and ghora, the original Indian form of kūda, is widely used as the name of the Knight. Chemōr (chemūr) is given as in colloquial Malay use only by Blagden (JRAS., 1898, xxx. 376). Bīdaq and the two technical terms sah and mat have been taken from the Arabic game. Marsden (op. cit.) gives the alternative terms māil (Malay, = dead) and tammat (Arabic, = finished) as also in use.

Of the Javan names, mantri is Sanskrit. So also is probably pateh (Skr. pati = lord or master). Ratu (= king), jaran (= horse), and prahu (= boat) are all Malay. From this it would seem that the Javan nomenclature preserves an older usage. On the other hand the disappearance of the Elephant and its replacement by two Counsellors is obviously the result of intercourse with the Dutch, with whom the corresponding piece has been long called by the name of Counsellor.¹⁰ It is more difficult to account for the replacement of the Chariot by the Boat. The same change has been made in the chess of the neighbouring lands of Siam and Annam, and also in the game as played in Bengal, where, however, it cannot be shown to be older than 1500. But Siam has exerted hardly any influence upon Malayan customs, and it is difficult to believe that Bengal can have had an influence sufficiently strong to affect the Javan nomenclature. I think it more likely that the change was made independently. The Chariot or Cart is of little use in a land of jungles, and it may very well have been replaced in chess by the Boat as representing the more usual means of transport.

Collectively the chessmen are called būwah chātor, i.e. the pieces (lit. fruit) of the chess.¹¹

At the commencement of the game the chessmen are arranged as in the Indian game (diagram, p.80) with the one exception that the relative positions of the Rāja and Mantri are reversed. In the Javan game, if MacGleans (Sch., 1867, 226) is correct, the Indian arrangement is followed. The Mantri is stationed at the right-hand of the Rāja. The arrangement is consequently crosswise. The powers of move of the pieces hardly differ at all from those which existed in European chess in the middle of the 16th century. The Mantri, Gājah, Kūda, and Tēr have exactly the same moves as their respective equivalents, the Queen, Bishop. Knight, and Rook, in modern European chess. The ordinary move of the Rāja is identical with the ordinary move of the King to any adjacent square. In addition he possesses certain powers of leaping into a square two squares distant. This liberty is not uniform throughout the Malay lands. In Borneo, according to Raja Brooke of Sarawak,

The King, when checked for the first time, has the right of making the Knight’s move, or to move two squares. After this sally, he is reduced to the same powers as a European King. The first move (in which he can of course take), on being checked, alters the game considerably, as one great object then becomes to prevent the check of your own King early in the game, and to gain a check of your adversary … for it will be evident if the King be once checked, he is deprived of one great advantage which your adversary still holds. Castling is not allowed except in two moves, the first being the Castle’s move up to the King, and on the King receiving a check, he can exercise his right of jumping to the inside of the Castle.¹²

In Java, according to Sir T. Stamford Raffles,

The King, if he has not been checked, may move two squares the first time, either as a Knight or otherwise…. The King cannot castle after having been checked. Castling is performed by two moves; the Castle must first be brought up to the King; after which the King may pass over the Castle at any future move, provided he shall not have been checked, or that no piece has occupied the square he would move into.¹³

According to Mr. H. O. Robinson,

Castling is effected in various ways in different parts of the Malay Peninsula and Straits Settlements; the recognized method in Selangor is to move two squares whether a piece intervenes or not,¹⁴ but not in conjunction with one of the Rooks. This is permitted even if the King is in check. The King may, also, before he is checked or moved from his own square, once move or take like a Knight. In Clifford and Swettenham’s Malay Dictionary it is stated that the King may, also, if he has not moved or been checked, move once over two vacant squares¹⁵; this privilege move is unknown to the Selangor Malays.

Finally v. Oefele says that in Batak chess the King may, for his first move, move from e1 to any of e2, e3, d1, d2, d3, e1, c2, f1, f2, f3, g1, g2—12 squares in all. Five of these are in virtue of its ordinary power of move, and 7 are to a second square. There are also two other squares, viz. c3 and g3, which are also only distant two squares, but no mention is made of them, and we must conclude therefore that the old leap of the Elephant in Arabic chess is prohibited. The leap may be made to remedy the first check, but at no subsequent turn of play, even if the first check is remedied by the interposition of a man or by the capture of the cheeking piece.

The use of the term ‘castling’ is of course inaccurate, since the manœuvre intended takes two moves. The leap naturally follows the Rook’s move, since the latter piece has no power of jumping. The manœuvre is quite well known, and occurs nine times in the nine games from native play that v. Oefele gives; on two of these occasions the King leaps out of check. In another game he makes the Knight’s leap in order to capture a Pawn.

The differences in practice are accordingly in connexion with two points: (1) whether the Rāja can or cannot make the leap when checked for the first time; and (2) to which of all the squares two steps distant the leap can be made. The rules given by the older observers are not sufficiently explicit here.

Every Bīdaq or Pawn is permitted the double-step for its first move, precisely as in European chess. Variety of practice appears to occur in connexion with taking in passing. Raja Brooke (Borneo) says:

A Pawn, moved out, cannot pass an adversary’s Pawn; his first move being restricted to one square in this case.

Sir T. Stamford Raffles (Java), on the contrary, allows the Pawn ‘passar battaglia’;

The Pawn may move two squares the first move, even though it should pass the check of an adversary’s Pawn.¹⁶

Robinson and v. Oefele give the rule thus: A Pawn can only take another Pawn in passing when its own advance is blocked by another Pawn; e.g. with white Pawns on g2 and h3 and a black Pawn on h4, if White plays Pg4, Black may reply P × P in passing¹⁷; if however there were no Pawn on h3, Black could not take the Pawn on g4 in passing, because he is not now blocked. This is a refinement which a casual observer would miss, and it is quite possible that it is the rule in Borneo and Java, and that the apparent discrepancy does not really exist.¹⁸ Robinson notes a further peculiarity in Pawn-play among the Selangor Malays. If White has a Pawn on h2, and Black a Pawn on g3, White being to play, he cannot play Ph3 or Ph4, but must play P × P, i.e. if he move the h-Pawn. If however White had also a Pawn on f2, he may now play either of the Pawns to its 3rd or take the Black Pawn, but he may not play either Pawn to its 4th.

V. Oefele states that the Bataks allow the King’s Pawn to defer its double step until its second move, e.g. 1 Pd3; 2 (or later) P(d3) d5; in such a case it is liable to be captured in passing on its second move, with similar conditions to those already given.

Pawn-promotion is quite different from the European practice. Generally a Pawn is promoted immediately on reaching the 8th rank only on the corner squares. Elsewhere it has to make some further move or moves. Raja Brooke says ‘two extra moves’ and illustrates the rule in the case of a Pawn played to c8; it is promoted by 1 Pb8, 2 Pa8; or 1 Pb8, 2 c7 or a7; or 1 d8, 2 e8 or e7. He explains that ‘this is a delay rather than a prevention, as from the number of squares which may be taken, it is extremely difficult to guard them all’. Sir T. Raffles, on the other hand, says that the Pawn after reaching its 8th rank on any file excepting the Rook’s files ‘must retrograde three moves before it can become a Queen’. This in Zimmermann’s somewhat loose description becomes: ‘the 3 joy-leaps (Freudensprünge) of Ströbeck are necessary before queening a Pawn.’ Robinson’s full account will again help to clear up these discrepancies. He says:

When a Pawn has reached the eighth square on the Rook’s file it queens at once; the player has also the option of selecting any other piece. If on reaching R7 a piece on Kt square is en prise and captured on the next move, the Pawn must move back one square diagonally before queening. On reaching the eighth rank of the Knight’s file it has to move back one square diagonally, either to the right or left, before queening; on the Bishop’s file two squares, and on the King’s or Queen’s file three squares.

I think it is obvious from this that the two diagonal lines that are drawn on the chessboard of the Peninsula are associated with this rule of promotion. The diagonals pass through the Rook’s squares, and promotion takes place at once, the Kt square is distant one square diagonally, and an additional diagonal move is necessary before promotion takes place. The B square is distant two, and the K and Q squares are distant three squares, and in these cases two and three diagonal moves are respectively necessary.

V. Oefele’s rules of Pawn-promotion are different again. Some of the Bataks do not know any rule, and when a Pawn has reached its eighth rank it turns about and retraces its way square by square across the board still moving and capturing as a Pawn, and it has the possibility of marching up and down the board an unlimited number of times.¹⁹ Generally the Batak players require an additional diagonal move to be made before promotion is possible. The two concluding steps—that from the 7th to the 8th rank, and the diagonal step—may both be made in the same turn of play, a double move called gelong, which is subject to the opponent’s right to take the Pawn in passing on the 8th rank. A Pawn may make a capture on the second move of the gelong. The gelong may not be played if the Pawn give check by the first part of the move. For example: White P on e7, Black R on f7. White can play P–e8–f7 taking the R, all in one move. If, however, the Black K be on d7 he can only play Pe8, check! Similarly, if it is possible to take a piece on the 8th rank, this capture is obligatory when the Pawn is moved, and the gelong is forbidden. In these two cases a second move is necessary to secure the right to promotion. Apparently the promotion is still incomplete and the Rāja must next make a move.²⁰ The promoted Pawn is now permitted to move in accordance with its new dignity, but it is still debarred from making a capture until its second move. It is not stated whether the Pawn is immune from capture during these operations. When finally promoted it can take the rank of any piece.²¹

This is a very long and complicated process, and very different from the rule as given by Robinson. It will be remembered that the Batak board is also covered with an elaborate network of lines which would not suggest a rule for promotion in the same way as the board of the mainland.

It is usual on the mainland to warn a player that his Mantri is under attack. Blagden gives mor as the call for this purpose. Robinson gives dōman as used at Selangor when the capture is threatened by another Mantri, and mā as used when any other piece makes the threat.

The term for discovered check is aras.²² This is derived from the Arabic i‘ra (Per. ‘irā, Hindustani ‘arop) which is regularly used in this sense in the earlier writers. Robinson gives aras sah as meaning double check and aras mā as a check which forks the Mantri. V. Oefele notes that the Bataks make a distinction between sah, direct check, and aras, discovered check. If the latter is irremediable—i.e. in European parlance is mate—the Batak calls the game drawn (sri): e.g. White, Kg1, Qh6, Kth7; Black, Kh8; the move Kt(h7)f6 is aras and the game is drawn. This leads to a still greater anomaly, a piece which is covering a check is deemed to have no power of giving check to the opposing King: e.g. White, Kg5, Bg4; Black, Ke2, Rf3, Pd3; White can calmly play Kf4 and draw the game.

Stalemate, called metuh (v. Oefele), or mūttu (Robinson), is reckoned as a draw.

There appear to be no special rules respecting Bare King in the Batak game. Sir T. Raffles says for Java:

A piece or Pawn must remain on the board till the last; if the King is left alone it is considered as stalemate, and he wins.

The allusion is probably to the English rule of stalemate at the beginning of the 19th century, in which the King who was put into the position of stalemate was counted as having won the game. MacGleans (Sch. 1867, 227) says of Java, however, that Bare King is a drawn ending. At Selangor the rule is different again; Mr. Robinson says:

Towards the end of a game care must be exercised in not capturing all the opponent’s pieces, for if the King be left solus the game is practically drawn, as he may move just as he pleases, like a King, Queen, Bishop, Knight, Rook or Pawn! He is then termed Rāja Lela with powers to bermaharaja lela, i.e. to play the Maharaja Lela.²³ …

The fact that the game is generally played for a stake naturally leads to the game at odds being often played. V. Oefele notes that the usual odds given by a strong player is to undertake to mate the opponent on one of the four central squares (d4, e4, d5, e5). This is called Tepong.

The crosswise arrangement of the Rājas, combined with the modern powers of move, has led to the prevalence of the wing attack in the actual game. Raja Brooke remarks that the ordinary method of opening the game in Borneo was to advance the QRP, the QKtP and QBP and to manœuvre the Q behind them. This is well illustrated in the nine games which v. Oefele gives from Batak play. After recording some games played by natives in his neighbourhood, he arranged a match between the best of the local players, by name Singambati, and Sibayak, whom popular opinion declared to be one of the best living Batak players. Sibayak had no difficulty in beating his opponent by 4–0. From his experience of native play, v. Oefele states that there are certain regular lines of opening play which are popular among players. The better players observe the rule that a piece once touched must be played.²⁴ The study of the simpler endings is also attempted with some system.

I select three games from v. Oefele’s work as illustrating well the main features of Batak play. In all of them I give the move to White, and the King’s are to be placed upon d1 and e8.

MALAY CHESSMEN (SELANGOR)

From the Skeat Collection in the Museum of Archaeology and Ethnology, Cambridge

There is no systematic study of the problem in Malay chess, but v. Oefele notes that a position is occasionally arranged on the board and a wager laid upon its solution. One such position that he had seen is the well-known European problem, White, Ke5, Re1; Black, Ke8, in which White gives mate on the third move.³¹

The chessmen in use on the mainland are generally clumsily carved from soft wood, with no distinction of colour, the one side being only distinguished from the other by a daub of lime or paint. Mr. Skeat tells me that the Pawns are often made afresh on each occasion of play. Ivory sets for royal use, and other sets of harder wood are not unknown. I give illustrations of some of the chessmen in the Skeat Ethnological Collection, and of some other sets as well. The more highly finished chessmen approximate in pattern to the modern Muslim pieces used in India. Since the Malays of the Peninsula are Sunnite Muhammadans of the Shafi‘ite school, the use of carved pieces, images of the actual forms represented, is forbidden by their religion.

In Sumatra, it is usual to make fresh chessmen on each occasion of playing. This only occupies about 10 minutes. A piece of bamboo or the midrib of a palm leaf is obtained and the pieces are quickly cut after a conventional pattern. The two sides are distinguished by slight variations in the shape. The pattern does not look to me to be derived from the Muslim type of the mainland. Most noteworthy is the fact that the Mantri is made the tallest of the pieces. The Kūda, with head cut aslope, may be a recollection of an early type of European Knight which is still occasionally repeated in English sets, and the Tēr with its cleft in the top recalls the old shape of the European Rook.

The country whence the Malays obtained their chess has been represented by different writers as Arabia, Persia, and India.³² The philological evidence derived from the nomenclature is only satisfied by the hypothesis of an Indian ancestry, with later modifications as a result of the knowledge of Arabic which resulted from the introduction of the Muslim religion from Southern India. It is not improbable that the Tamil and Telugu terms were also introduced with Muhammadanism.

I. A set of hard wood in the possession of Mr. Robinson.

II. From v. Oefele.

III. From Mr. Claine’s paper, BCM. 1891.

MALAY CHESSMEN.

The evidence of the practical game points to Southern Europe³³, and suggests that extensive modifications have been made in rule and move as a result of the intercourse with Portuguese and Dutch since 1500. The existing variations all appear to me consistent with the view that the European practice of the middle and later half of the 16th century remodelled the native game. The differences are superimposed, not fundamental. They occur just in those points in which uncertainty exists to-day among beginners, or in circles out of touch with the literature of the modern game. At the same time the use of the unchequered board, and the whole question of Pawn-promotion, is still pure Asiatic. To the objection that the European powers of move had already taken root in India, and that there is the simpler possibility that the change came via India, the Pawn’s move seems a sufficient answer. Had the change come from Southern India, we should have found the double step restricted to particular Pawns, or hedged about with conditions: we should probably also have found restrictions placed upon the free promotion to the rank of any piece.

Table of Contents

Chess not a classical game.—The name zatrikion.—First references in Arabic works.—References in late Greek literature.—Ecclesiastical censures.—Chess in the Turkish rule, and in modern Greece.

It was a common belief among mediaeval writers that chess was a game, if not of Trojan, at least of Greek invention, and that various references to the practice of games among the Greeks and Romans in classical times related to chess. In the light of the facts of the history of the spread of chess which have been narrated in the preceding chapters, this view can no longer be seriously maintained. Quite apart from the fatal anachronisms involved in the claim, it can be shown to be improbable, if not impossible, on a priori grounds, from an examination of the character of the references and allusions to board-games in classical Greek. When these references are carefully examined, it is found that they reveal not the slightest trace of any allusion to any characteristic at all similar to the essential characteristics of chess—pregnant in possibilities of allusion, simile, or metaphor as these have proved in every chess-playing country. It is inconceivable that such a silence could have existed throughout Greek and Latin literature had any of the classical games shown those peculiarities of piece, form, and move which are the special property of chess. Nor, again, would it have been necessary, as v. d. Lasa has acutely pointed out, for the Byzantines to have introduced a new name for chess if the petteia, or the game of the sacred way, or any other of the classical games had been chess. Slight and conjectural as is our knowledge of these games, whether requiring the agency of dice or not, this much at least is certain: none of them was chess and none of them was like chess.¹ Games of skill some of them certainly were, but all lacked the vitality that chess has always shown, and it is clear that they had dropped into desuetude by the sixth century of our era, for long before that date commentators were revealing, by their curious and inconclusive attempts to explain the classical allusions to the petteia and other games, their complete ignorance on the subject.

With the games of the Byzantine period (A.D. 365–1450) we are not much better off for information. Our knowledge is small and goes but little beyond the names of a few of the games that were current. Our want of knowledge may, it is true, be due in part to the uninviting nature of the later Greek literature. The number of scholars who have ventured upon that dreary and unprofitable field is very small, and we are practically indebted for what little we know of the Byzantine games to the first zeal of the scholars of the Renascence; no later writer has added anything of material value to the information first arranged by the four scholars of the 17th c., Jules-César Boulenger, Johannes Meursius, Daniel Souterus, and Andrew Senftleben, the salient facts of which may be seen most conveniently in the pages of Gronovius or the lexicons of Ducange.²

That chess should be found among the games of the later half of the Byzantine period is not surprising. On the contrary, when the political intercourse which subsisted between the Eastern Empire and the later Sāsānian monarchy and the ‘Abbāsid caliphate is remembered, together with the general adoption of Persian customs and luxuries at the court of Constantinople, it would be strange indeed if a knowledge of chess had not penetrated to the Imperial court.

The earliest references to Eastern games in Byzantine Greek are probably those relating to tabla, in later Gk. taula, which was probably identical with the Persian and Arabic nard or nardshīr. Etymologically the word tabla is merely an adaptation of the L. tabula, table, which was already used by Juvenal in the sense of gaming-table, and at a later time appears to have become the ordinary name for the ludus duodecim scriptorum of the classical period.³ If this game was ever played in the Eastern Empire, it was soon supplanted by the Persian nard, a game of the same class, and the name of tabla was transferred to this latter game.⁴ It is this game tabla which is mentioned in some epigrams of Agathias the scholastic of Myrine in Asia, who flourished A.D. 527–67; the longest of these (Anthol Pal., IX. 482) describes an extraordinary position in the game which had occurred to the Emperor Zeno (A.D. 475–81). The position has been recovered independently by M. Becq de Fouquières and by Prof. Jackson of Cambridge, and their reconstruction shows that the game was identical with the Persian nard. Hyde (ii. 255–6) quotes notes on the Gk. tabla from Cedrenus, Suidas, and Isaac Porphyrogenitus which contain the germ of the astronomical explanation of nard which we have met already in the Chatrang-nāmak. It is noteworthy that this Greek name for nard has replaced the older name in Syria, Turkey, and generally along the S. coast of Asia, where the game on the backgammon board is now commonly called tawūla.

Chess makes its appearance in Byzantine Greek under the name ζατρíκιον, zatrikion. This word is unknown in classical Greek, and is incapable of explanation from native roots. As Hyde and Forbes have shown, the word is ‘simply a barbaric or foreign word with a Hellenic termination’. It can be shown that this form answers exactly to the Middle Persian chatrang, when allowance is made for the different range of the Greek and Persian alphabets. ‘The Greek alphabet’, writes Forbes, ‘had no letter or combination of letters capable of expressing the sound of the Persian ch-, and as the nearest approximation they employed for that purpose the letter ζ, z.’ For similar reasons they had to transliterate the Semitic sh-, by σ, s, or by σι, si-. The nearest Greek approximation to Per. chatrang would be ζατρáγκ or ζατρέγκ, and this, on Greek analogies, gave ζατρíκιον, the form actually found. (An n sound in such a position was often transposed or altogether suppressed.) Shaṭranj, the Arabic and modern Persian name of chess, would have given satrantz. Ducas has σαντράτζ (with n transposed).

The form zatrikion accordingly becomes of importance in connexion with the date of the introduction of chess into the Byzantine Empire. The presumption is that the knowledge of the game was obtained at a time when the Persians still used the older form chatrang, and not from the later Persians, the Arabs, or the Syrians, all of whom had substituted the form shaṭranj for it. Forbes (190–5) assumed that this required the introduction to be anterior to the Muhammadan conquest of Persia, i.e. before the middle of the 7th cent., and fixed upon the exile of Khusraw (II) Parwīz as the date which with ‘strong possibility’ saw the introduction of chess into Eastern Europe. That would place it in the first quarter of the 7th century.

Forbes, however, assumed that the influx of Arabic words and forms into Persian was an immediate result of the Islamic conquest, A.D. 638–51. Such was certainly not the case. Chatrang may have easily remained in use for another 200 years, the earliest evidence for its disappearance belongs to the 3rd century of Islam (A.D. 830–930). All we can assert is that the philological evidence points to the introduction of the word zatrikion not later than the 9th century A.D., while it does not at all necessarily follow that the practice of the game began so early: the knowledge of the existence of a thing may precede its use by a considerable interval of time. It is quite possible that the word zatrikion came into Greek first in accounts of travel in Persia, or in descriptions of Persian life.

Sound as these conclusions undoubtedly are, they cannot be substantiated by contemporary Greek records, and not one of the earlier uses of the word zatrikion can be dated with any approach to exactness. The earliest evidence exists only in Arabic works, and establishes a knowledge of chess and its technicalities at Byzantium by the year A.D. 800. In the K. akhbār ar-rusul wal-mulūk of the historian at-Tabarī (B. 224/838, D. 310/923)⁵ we read:

It is related that when Niqfūr (Gk. Nicephorus) was king, and the Byzantines had assembled in allegiance to him, he wrote to ar-Rashīd: ‘From Niqfūr, King of Byzantium, to Hārūn, King of the Arabs, now the Empress to whom I have succeeded estimated you as of the rank of the Rook, and estimated herself as of the rank of the Pawn, and paid a tribute to you, which you rightly should have paid to her. But this was because of a woman’s weakness and folly. When therefore you have read my letter, return the tribute that has been previously paid to you, and come yourself with what you have to repay. If not, the sword is between us and you.’ It is reported that when ar-Rashīd read this letter, his wrath was kindled … and he called for an ink-pot and wrote on the back of the letter: ‘In the name of God! the compassionate and merciful! From Hārūn, Commander of the Faithful, to Niqfūr the dog of Byzantium. I have read your letter, son of an infidel woman. The answer is what you will see, not what you will hear.’ And he struck his camp that day, and marched until he encamped at the gates of Hiraqla (Gk. Heracleia, 65 m. N.W. of Tarsus).

The ruthless conduct of this invasion soon compelled Nicephorus to consent to continue the tribute that his predecessor Irene had paid. The incident is told under the year A.H. 187 (= A.D. 802), in which Nicephorus became Emperor.

The rather later geographer and historian al-Maṣ‘ūdī (D. 345–6/956) refers to the Greeks in connexion with chess in two places in his Murūj adh-dhahab. At the close of his account of the invention of chess in India in the reign of the mythical King Balhait, he says:

The Greeks (al-Yūnānīyan) and Byzantines (ar-Rūm) and other peoples have special theories and methods about this game, as we may see in the works of chessplayers from the most ancient down to al-‘Adlī and aṣ-Ṣūlī

And much to the same effect at the conclusion of a digression on the modifications of chess (among which is ‘the round board attributed to the Byzantines’) he remarks:

The Indians and others, the Greeks, Persians, and Byzantines who play at chess have given accounts of the manner and fashion of the pieces in chess, its arrangements, its beginnings, the various motives underlying it, its peculiarities, and the classifications of the qawā’īm and mufridāt, and the classes of the noteworthy manṣūbāt.⁶

Greek literature and tradition are alike silent as to the existence or otherwise of these works and theories, and when we turn to the Arabic chess MSS. which are based upon the works of al-‘Adlī and aṣ-Ṣūlī, we find the only references to Greek chess relate to the philosophers Hippocrates and Galen, and to Aristotle. Hippocrates and Galen apparently found in chess a potent antidote to diarrhoea and erysipelas, and prescribed it with success, while Aristotle figures among the many hypothetical inventors of chess. Another story tells how Galen once met a friend whom he had not seen for some time, and learnt that he had been into the country to see a farm which he had purchased with the result of his gains at chess, whereupon the physician exclaimed with what sounds like a strong flavour of irony, ‘What a fine thing chess is, and how profitable!’ Pure fiction, the whole of it, of course.

Most of the MSS. agree with al-Maṣ‘ūdī in giving some account of round chess under the title of ash-shaṭranj ar-rūmīya, or Byzantine chess, while they lay stress upon the fact that it is only a modification of the ordinary or Indian chess,⁷ It is difficult to understand its designation unless there were some historic justification for it.

It would appear that the earliest use of the word zatrikion occurs in works treating of the interpretation of dreams. This is a Science which was apparently first exploited by the Greeks,⁸ but soon passed to Persian and Arabic writers. The Muhammadan tābι‘Muḥammad b. Sīrīn (B. 33/653–4, D. 110/729), of Persian parentage, was skilled in this lore, and became the first of a long line of Oriental writers on the subject. One of these Arabic works was retranslated into Greek, and thence into Latin by Leo Tuscus in 1160. A later Latin version is due to the German traveller John Leunclavius (B. 1533, D. 1593), who ascribed the Greek work to Apomazares, in whose name we may recognize the Arabic oneirocritic Abū Ma‘shar (D. 272/885). Nicholas Rigault (B. 1577, D. 1654) printed the Greek text in 1603 with Leunclavius’s translation, and ascribed it to Achmet fil. Seirem. This is generally understood to mean Muḥammad b. Sīrīn, though on the strength of the Greek version Achmes appears in some lists of Greek authors as flourishing, now as early as A.D. 750, now as late as A.D. 950! Since the work contains the interpretation of a dream that happened to al-Ma’mūn, who reigned A.D. 813–833, it cannot be b. Sīrīn’s work, and Bland has shown⁹ that there are grounds for believing that it is of Christian authorship. The Greek can hardly be earlier than the 10th century. Chapter 241 treats ‘Of zatrikion. From the Persians and Egyptians’.¹⁰

If any one dreams that he plays chess (zatrikizo, vb.) with a man he knows, they will quarrel over money affairs, &c.

If a king or grandee or general dreams that he plays chess, he will think of the place for joining battle with the enemy, &c.

If he dreams that he takes many pieces in the game,¹¹ he will take many of the enemy, &c.

If a king or grandee or general dreams that he has lost or broken or been deprived of his zatrikion, he will lose his army, &c.

Besides this passage, Ducange quotes two other references in MSS. accessible in his time, one attributed—but certainly wrongly—to Astram-psychus, in which twice occur the words ‘chess and tables’,¹² the other from an anonymous MS. on Persia, De arte Persica, ‘slaves and games of bolgon and chess and love of women.’ Neither of these passages can be dated, and the present location of the MSS. is unknown to me. The only point of importance about either appears to be that chess is associated with other notorious features of Persian luxury. It has probably never been in worse company.

A fourth instance occurs in a scholiast’s commentary on Theocritus, Idýll, vi. 18,¹³ where there is an allusion to the Greek game of petteia—‘he moves away the pebble from the line’ This, the commentator explains, ‘is a figurative expression borrowed from the phraseology of those who play at the game commonly called zatrikion’—an absurdity that provoked Dr. Hyde’s scornful comment, ‘quantum hallucinatus est Scholiastes!’ Here again we have no clue to the date of the writer.¹⁴

It is not until we come to the 12th century that we have an instance of zatrikion to which we can assign a definite date. In the twelfth book of the Alexiad of the Princess Anna Comnena (D. 1148), a laudatory biography of her father, the Emperor Alexis Comnena (D. 1118), we read in an account of the Emperor’s recreations:

He had certain familiar friends with whom he played chess, a game that was discovered in the luxury of the Assyrians, and was brought to us.¹⁵

Here again chess is associated with Oriental luxury. Assyria, of course, was no longer a kingdom in Anna Comnena’s day, and her use of the name probably only refers to the traditional splendour of the earlier Oriental empires.

The Emperor Alexis’s fondness for chess may have been the cause of raising a powerful and bitter hostility to the game. It is at least singular to find that the first ecclesiastical denunciation of chess on the part of the Eastern Church was voiced by John Zonares, who, after filling the post of commander of the Emperor’s bodyguard, retired as a monk to the Monastery of Mt. Athos and died there in 1118. It was during his retirement that he wrote, his commentary on the canons of the Eastern Church.

That the early mediaeval Church viewed the use of dice with strong disfavour is evident from the attempts that were made to suppress it by legislation. The early list of rules known as the Apostolic Canons¹⁶ requires both clergy and laity to give up the use of dice.

42. A Bishop, Priest, or Deacon addicted to dice (Gk. κύßoι, Lat. alea) shall either give them up, or be deposed.

43. A Sub-deacon, Reader, or Singer doing the same shall either give them up, or be deposed. So also the laity.

These rules were adopted by the Trullan Synod (Third Council of Constantinople) in 680,¹⁷ and have since then formed part of the Nomo-canon of the Eastern Church. It was natural that, in the course of time, the attempt should be made to explain the prohibition of kuboi or alea by defining exactly the games which were to be included under these terms. This attempt was not confined to the Eastern Church: the later Latin use of alea as the name of a game helped to confuse the lawyers of the Western Church, and we shall find Cardinal Damiani arguing in a letter of 1061 that the prohibition of alea extended to chess. The Western Church took this view for a considerable time.

Zonares makes the following note on the 42nd rule of the Apostolic Canons:

Because there are some of the Bishops and clergy who depart from virtue and play chess (zatrikion) or dice or drink to excess, the Rule commands that such shall cease to do so or be excluded; and if a Bishop or elder or deacon or subdeacon or reader or singer do not cease so to do, he shall be cast out: and if laymen be given to chess-playing and drunkenness they shall be excluded….

We shall see later that this extension of the term kuboi was for long adopted by the Russian Church, and we may probably account in part for the paucity of references to chess in the Eastern Empire as being due to the intolerance of the Church.

The beginning of the 13th century saw the Latin or Western Emperors established in Constantinople, who must have known chess in Western Europe before they laid hands on the Empire of the East. The result of this may be detected in the latest reference to Zatrikion in Byzantine Greek. The history of Ducas, written about 1400, nearly at the close of the Eastern Empire, contains an account of the incidents which led to the naming of Tīmūr’s son Shāhrukh from a technicality of chess. In this passage Ducas adds the information that the Persians call zatrikion santratz (σαντρàτζ), and the Latins call it scacum (σκáκον). Later on he uses σκáκον for a chessman and σκáκον πατγνíον for the game of chess, which are evidently adaptations from the Latin scacus, a chessman, and scacorum ludus, the game of chess. Shāhrukh is transliterated Siachrouch (σιαχρούχ), with the information that the Latins call it σκáκω ζóγκω, a curious misrendering of scac-roc. It seems clear to me that Ducas knew more of the Latin than of the Greek chess.

There is one branch of the later Greek literature, fairly circumscribed in extent, which might possibly give us some reference to chess earlier in date than any I have cited. The mathematical problem known as ‘the doubling of the squares of the chessboard’ may have been known to the later Greek mathematicians, as we find it included in the oldest Western mediaeval MSS. on mathematics. The Greek MSS. have not so far been examined for this purpose.

With the fall of Constantinople (1453) and the last outposts of independent Christianity in Asia Minor (1461), the last vestiges of the Byzantine Empire passed away, and the Greeks became a subject race and largely adopted the language of their conquerors. The game of zatrikion, whatever special points and rules it may have possessed, must be held to have become obsolete, and its very name soon passed into oblivion. Whatever chess was played would assuredly be the Turkish chess of the ruling race. A curious confirmation of this at the very end of the Turkish dominion over Greece itself is to be found in the name, ‘the Greek Defence’, which Allgaier, following the usage of Viennese chess-players, gave to the Fianchetto defences, which are still to-day a striking characteristic of the native Turkish chess. This result was probably assisted both by the small degree of popularity that chess would seem to have secured among the Greeks,¹⁸ and by the ecclesiastical opposition to its practice.

With zatrikion forgotten, it is only natural to discover the use of a new name more closely representing the Turkish shaṭranj. The poverty of the Greek alphabet necessitated changes in the form of this word when the attempt was made to reproduce it in a Greek dress. The Semitic sh was variously replaced by s or si, the j by tz or z. Shaṭranj accordingly gives rise regularly to santratz, as in Ducas, or santraz, the form which Hyde gives as in use in his time.¹⁹ Modern dictionaries give santratsi as in vulgar use, and add still another form, Satrengion (Σατρένγιον, σατρέγγιον), which is a modern adaptation from the Egyptian dialect of Arabic.²⁰

Turkish chess has met the same fate in Greece that befell zatrikion, and the modern Greek has turned to the West for his knowledge of chess, and the name of the game, skaki (σκáκι), and the translations from the French which do duty for the names of the chessmen, betray at once the origin of the modern Greek chess of our day. The attempt to revive the word zatrikion, as seen in the title of the only Greek work on the game, the Encheiridion Zatrikion of Leo Olivier (Athens, 1894), is due to the workings of national aspirations.

Table of Contents

The chess works mentioned in the Fihrist, and other bibliographies.—MSS. used for the present work.—Other MSS. in European libraries.—Poems and impromptus on chess, &c.

The beginnings of the vast literature of chess are to be found in the Golden Age of Arabic, the first two centuries of the ‘Abbāsid caliphate, that short period during which alone Islam has shown any powers of original thought and discovery. In b. Isḥāq an-Nadīm’s great bibliographical work, the K. al-fihrist, compiled 377/988, we find a section devoted to the authors of books on chess.

These are the chess-players who wrote books on chess.

AL-‘ADLĪ. His name is (left blank). He wrote Kitāb ash-shaṭranj (Book of the chess). He also wrote Kitāb an-nard (Book of the nard).

AR-RĀZĪ. His name is (left blank). He was of equal strength with al-‘Adlī. They used to play together before Mutawakkil (Caliph, 233/847 – 248/862). The book Latīf fī’sh-shaṭranj (Elegance in chess) is by him.

AṢ-ṢULĪ. Abū-Bakr Muḥammad b. Yaḥyā, who has been mentioned already. He wrote Kitāb ash-shaṭranj, the first work, and Kitāb ash-shaṭranj, the second work.

AL-LAJLĀJ. Abū’l-Faraj Muḥammad b. ‘Obaidallāh. I have seen him. He went to Shīrāz to the king ‘Adudaddaula (ruled 338/949 – 366/976), and died there in the year 360/970 and a few. He was excellent at the game, and among the books on it Kitāb manṣūbāt ash-shaṭranj (Book of chess-positions or problems) belongs to him.

B. ALIQLĪDISĪ. Abū Isḥāq Ibrāhīm b. Muḥammad Ṣāliḥ. He is reckoned among the brilliant players, and wrote the Kitāb majmu‘fī manṣūbāt ash-shaṭranj (Collection of chess problems).

The other much later great Arabic bibliography, the Kashf aẓ-ẓunūn fī asāmī’l-kutub wal funūn of Ḥājjī Khalīfa (D. 1068/1658) has a shorter catalogue of chess books.

10224. Kitāb ash-shaṭranj by the authors Abū’l–‘Abbās Aḥmad b. Muḥammad as-Sarakhsī, the physician, who died in the year 286/899; Yaḥyā b. Muḥammad aṣ-Ṣūlī; and a later author who wrote in Persian and boasts not without arrogance that he was the best player of that game in our own time in the whole world. He drew the figure of the chessboard and sketched the pieces and enumerated the authors who had previously written on this game.

As-Sarakhsī¹ ranks as the most important of the pupils of the Arabic philosopher al-Kindī, who lived in Basra and Baghdād in the caliphates of al-Ma’mūn and al-Mu‘taṣim. He himself held a position at the court of al-Mu‘taḍid, but fell into disfavour by revealing a secret which this caliph had entrusted to him, and was thrown into prison and executed, 286/899. An-Nadīm, however, makes no mention of a chess work in his list of as-Sarakhsī’s writings,² nor does al-Qiftī (568/1172 – 646/1248); a later biographer of as-Sarakhsī, b. Abī Uṣaibi‘a (B. 600/1203, D. 668/1270), on the other hand, who wrote on the lives of the Arabic physicians, mentions it under the title K. fī’sh-shaṭranj al-‘ālīya (Book of the higher chess) in his K. ‘uyūn al-anbā’ fī ṭabaqāt al-aṭibbā’.

An-Nadīm left a blank in the place of al-‘Adlī’s personal name, thereby implying that he was unable to discover it. In some modern works, e.g. in the catalogue of the Library of the Sultan ‘Abd-al-Ḥamīd Khān, however, his name is given as Abū’l-‘Abbās Aḥmad al-‘Adlī, thus making his personal name identical with that of as-Sarakhsī. I have been unable to discover the authority for the modern statement, and am inclined to think that it has arisen from the assumption that al-‘Adlī and as-Sarakhsī were one and the same person. This assumption would certainly account for the omission of al-‘Adlī’s work in Hājjī Khalīfa’s bibliography, but it introduces chronological difficulties. We know from aṣ-Ṣūlī that al-‘Adlī had stood alone in the first class of chess-players for some considerable time when he was defeated by ar-Rāzī in a match which we know from an-Nadīm was played in the presence of the caliph Mutawakkil (A.D. 847–862). After his death ar-Rāzī in his turn stood alone in the first class for some time, and was dead before aṣ-Ṣūlī came to the front under al-Muktafī (A.D. 902–8). It seems reasonable to infer that al-‘Adlī was past his prime at the time of his defeat, and that he probably did not survive it many years. As-Sarakhsī, on the other hand, must have been still a young man in Mutawakkil’s time, since his master al-Kindī flourished A.D. 813–842, and he himself only met with his death so late as A.D. 899. Moreover, the MS. RAS gives al-‘Adlī the local epithet of ar-Rūmī, which implies that he was a native of some town in the lands of the old Byzantine Empire. Had he come from Sarakhs, al-Khurāsānī would have been the more appropriate designation. On the other hand, if, as seems most likely, the two men were really distinct, the silence of all the Muslim chess writers concerning as-Sarakhsī is somewhat remarkable.

Of the other authors named above, ar-Rāzī³ has been identified with the celebrated physician Muḥammad b. Zakarīyā ar-Rāzī, the ‘Rhasis’ of mediaeval science, who died 311/923 or 318/932. This identification is palpably false. The chess-player belonged to an earlier generation and was dead before A.D. 900. Of aṣ-Ṣūlī and al-Lajlāj I shall have more to say in the following chapters; b. Aliqlīdisī is not otherwise known, but Hājjī Khalīfa’s anonymous and bombastic Persian MS. appears to be the one I refer to below as RAS.

B. ‘Arabshāh, the biographer of the great Tīmūr, in his digression upon the chess-players of the Court incidentally refers to another work by a contemporary of Tīmūr (D. 1405).

‘Alā’addīn Tabrīzī, commonly called ‘Alī ash-Shaṭranjī, has composed a treatise on the game of chess and its situations.

Finally Ahlwardt, in his Catalogue of the Arabic MSS. of the Royal Library at Berlin, gives without stating the source of his information the following list of chess works at the conclusion of his description of the chess treatises in the library.

Kitāb shaṭranj by Aḥmad b. Muḥammad as-Sarakhsī (D. 287/899). ’Istaw‘a’n-nahj fī taḥrīm al-la‘b bī’sh-shaṭranj, by Muḥammad b.‘Ali b. Muḥammad al-Hadhāmī b. an-Najār (D. 723/1323). Īqāẓ an-nāṣib fīma fī’sh-shaṭranj min al-manāṣib, by ‘Alī b. Muḥammad al-Mauṣilī b. ad-Duraihim (D. 762/1361).

I now come to the Arabic and Persian works of which I have been able to make use for this book.

There is much similarity about the MSS. which deal with the practical game, and it will be more convenient to summarize their contents in a table and so to avoid considerable repetition. There is usually an introductory section dealing with the legendary accounts of the invention of chess, and the evidence for the lawfulness of chess-playing for Muslims. Chapters dealing with the classification of players, with the relative value of the pieces, with the symbolism of the game, with the decisions as to the result in the simpler Endings of the game, with notation, and the derived chess-games generally follow. There are also chapters dealing with the normal positions for Opening play (the ta‘bīyāt), and the body of the work is devoted to a collection of manṣūbāt or problems. Less frequently we find an anthology of chess poems as a crown to the book. These MSS., it will be obvious, deal with nearly every aspect of chess.

1. AH = MS. ‘Abd-al-Ḥamīd I, Constantinople, no. 560.

2. C = MS. Khedivial Lib., Cairo, Muṣṭafa Pasha, no. 8201.

These are two MSS. of the same Arabic work, the Kitāb ash-shaṭranj mimma’l-lafahu’l-‘Adlī waṣ-Ṣulī wa ghair-huma, ‘Book of the chess; extracts from the works of al-‘Adlī, aṣ-Ṣulī and others.’

AH is one of the Arabic MSS. the knowledge of whose existence we owe to Dr. Paul Schroeder. It is no. xviii of v. d. Linde’s list (Qst., 331 seq.). It is a beautifully executed paper MS. of 142 leaves, 27.8 cm. by 21.2, written in a careful nashkī hand by Abū Isḥāq Ibrāhīm b. al-Mubārak b. ‘Alī al-Mudbahhab a1 Baghdādī, 535/1140, as we learn from a note on f. 54 b. Both the main title-page and the subsidiary one on f. 55 a are richly coloured, the titles being in the kufic character upon a blue ground.

C is no. viii of v. d. Linde’s list (Qst., 21). It consists of 157 leaves, 26 cm. by 18. From the richly illuminated title-page, now unfortunately much faded, it is evident that this MS. formerly belonged to a Sultan of Egypt, whom a former librarian, Dr. W. Spitta, identified from considerations of handwriting and ornamentation with Qāitbāi (A.D. 1468–96). He dated the MS. itself c. 770/1370.

Neither MS. gives any information as to the name of the writer who put together this compilation. A later note on a blank page at the commencement of AH attributes the work to al-Lajlāj, but the fact that this MS. includes a short treatise on chess principles, naming al-Lajlāj as its author, makes it very improbable that this player was the author of the whole work. In the official catalogue of MSS. in Constantinople libraries it is described as ‘560. Risāla fi’sh-shaṭranj, one volume in Arabic, by Abū’ l-‘Abbās Aḥmad al-‘Adlī’—an entry due to the occurrence of al-‘Adlī’s name in the title of the MS.

Neither MS. is complete. There are gaps in AH between ff. 75 and 76 (the latter leaf beginning in the middle of a problem solution), and 139 and 140 (the poem on the former leaf is incomplete). Ff. 121–123 should be placed between ff. 129 and 130, and f. 21 between ff. 22 and 23. The disarrangement of the entries on ff. 25–29 goes back to a MS. lying behind AH.⁴ C is a copy of AH, or of a MS. derived from AH.⁵ It is not so extensive, the text on ff. 133 b–142 b of AH being missing. There are also gaps between ff. 5 and 6, 17 and 18, and 23 and 24. The leaves from f. 34 onwards are now in great confusion; none, however, is missing, and they can be arranged in their original order with the help of AH.

The introduction to AH and C shows that aṣ-Ṣūlī’s book was largely a critique on al-‘Adlī’s. It runs as follows:

In the name of God, the compassionate and merciful! There is no prosperity except through God! Al-‘Adlī gives several accounts of the invention of chess, which Abū-Bakr aṣ-Ṣūlī criticizes. We narrate some of what al-‘Adlī relates, with aṣ-Ṣūlī’s criticisms thereon, and also the problems which al-‘Adlī placed in his book, with aṣ-Sūlī’s criticisms and appreciations. We have also added some problems from aṣ-Ṣūlī’s book, and some from other authors, together with the traditions which aṣ-Ṣūlī collected on the lawfulness of chess-playing.

The compiler accordingly claims to treat his authorities with some discrimination, and generally makes it clear from whom he is quoting. Extracts from previous writers are commenced by the words qāla’l-‘Adlī or’Ṣ-ṣūlī, as the case may be, and are generally in the first person.

The earlier chapters in AH are unusually full and informing. There is also near the end (AH, ff. 133 b – 135 a; not in C) an important tadhkira or treatise on chess principles by al-Lajlāj, of which I give the substance in Ch. XIV. The extensive contents of this MS. make it one of our best authorities for Muslim chess.

3. BM = MS. British Museum, Arab. Add 7515 (Rich).

This is a quarto MS. on vellum of 132 leaves, which was completed 16 Jumādā II, 655/1257. It formed part of the library that Claudius J. Rich (B. 1787, D. 1820) collected while Resident at Baghdād in the service of the East India Company, and was bought by the Museum Trustees from his widow.

Forbes (74) represented this MS. as a copy of a work written between 1150 and 1250. The arrangement of the MS. does not bear out this view. It has all the appearance of a work planned upon a larger scale than was carried out, the gaps in which the writer filled in later without regard to their surroundings. There are leaves missing between ff. 7 and 8, 16 and 17, 27 and 28, 34 and 35.

There is nothing in the MS. to show the name of its author, but he has made liberal use of al-‘Adlī’s work, and quotes from al-Lajlāj with approval. Aṣ-Ṣūlī is on the whole ignored; the few extracts from his work, e.g. from his preface on f. 8b, are unacknowledged. The text to fourteen of the problems is identical with that in AH, and it is possible that the as-Sulī extracts may have been taken at second hand from a compilation like AH. The MS. is dedicated to a Prince whose name has been erased. Forbes identified him from the special titles and epithets used with one of the Ayyūbid dynasty who ruled over Egypt, A.D. 1193–1250, but I cannot reconcile what is left of the name of the Prince in the MS. with the name of any member of this house, and Cureton (Cat. Arab. MSS. in the Brit. Mus., ii. 351, No. 784) does not pretend to identify either the Prince or his dynasty. From the compiler’s knowledge of al-Lajlāj’s work, I should be inclined to believe that the MS. was compiled farther East than Egypt, and possibly in Persia.

The most noteworthy feature of the contents is the brief chapter on the Openings of the writer’s time, f. 11 a. I quote this original contribution to the history of the ta‘bīyāt in Chapter XIV.

The front page, f. 1 a, contains a number of entries in later hands. These consist of (a) a title, Kitāb ash-shaṭranj al-Baṣrī, ‘al-Baṣrī’s chess book’, which is a manifest error due to the fact that a quotation from al-Ḥasan al-Baṣrī (D. 110/728) stands at the top of f. 2b: (b) a title Kitāb fī’sh-shaṭranj wa mausūbāt-hi wa mulaḥ-hi, ‘Book of chess, its problems and subtleties’: (c) a note in an 18th-century hand giving the differences of move between the chess of the MS. and the chess of the writer’s day, which I quote below, p. 354: (d) a note in a 15th-century hand giving the sum of the doubling of the squares of the chessboard: and (e) a calculation of the same total in Turkish.

4. L = MS. As‘ad Efendī, Constantinople, No. 1858.

A MS. of 81 worm-eaten leaves, which was discovered by Schroeder (Qst., No. XVII, pp. 382–9). The binding bears the title Risāla al-Lajlāj fī bayān la‘b ash-shaṭranj (‘Al-Lajlāj’s treatise on the demonstration of the game of chess’), and the title-page that of Kitāb ash-shaṭranj ta‘alif Abī’ l-Muẓaffar b. Sa‘īd ‘urifa bi’l-Lajlāj, ‘Book of the chess, composed by Abū’l-Muẓaffar b. Sa‘īd who is known as al-Lajlāj’ (i.e. the stammerer). The MS. is undated, but may be as much as three centuries older than an entry on the title-page chronicling the fact that the Sultan Bāyazīd Khān gave the book to his chief butler, Yūsuf b. ‘Abdallāh, the first day of Shawwāl, 893/1487.

The MS. is a treatise on the practical game, and contains a full analysis of certain of the more popular openings, with the view of establishing the superiority of the Mujannah Opening. It is in consequence a work of prime importance for the history of the practical game: it is the only work on the subject prior to those of the first analysts of the modern European game, and, being the work of a master of the first rank, who expresses his own indebtedness to his own master, aṣ-Ṣūlī, the greatest of all the Muhammadan masters, we may safely regard it as recording the highest point of development reached in the whole history of the older chess. The MS. is incomplete at the end, where it breaks off to give a problem which al-Lajlāj had mentioned—though not in the present work—under the name ad-dūlābīya (the water-wheel). Leaf 9 is out of place, it should come between ff. 37 and 38.

5. AE = MS. As‘ad Efendī, Constantinople, No. 2866.

An undated, anonymous Persian MS. of 609 pages, with the title Kitāb ash-shaṭranj, which is No. XXI of v. d. Linde’s list (Qst., 333). V. d. Linde gives no account of the MS., but merely quotes the opinion of Aḥmad Ḥamdī Efendī, a Turkish scholar who examined it for Schroeder, that it was a work of ‘no value’. This hasty judgement cannot be accepted. The MS. proves on examination to be a compilation treating of all branches of chess. The writer, however, has carefully excluded all reference to his sources, and only names ‘Adlī and Lajlāj Shaṭranjī as supporting certain verdicts in the Endgame. After a lengthy preface on the creation, of which the noblest work was man, and on man’s glory, to wit his intellect, of which chess and nard are the most striking fruits, the work continues with a close and complete translation of al-Lajlāj’s Arabic work which we possess in L. The leaves are in some confusion, but the text affords a valuable means of testing the accuracy of L, especially as AE contains 60 diagrams showing the position at various points of the analysis. It also supplies the conclusion which is missing in L.

The second section of the MS. consists of a long list of decisions on the Endings. The third section is an extremely valuable collection of 194 problems, with which I deal in Chapter XV.

6. V = MS. Vefa (‘Atīq Efendī), Eyyub, No. 2234.

A paper MS. of 77 leaves, 24 cm. by 19, one of those discovered by Dr. Schroeder, and No. XIX of v. d. Linde’s list (Qst., 390–6). Schroeder, gave as its title Manṣūbāt li Abī Zakarīyā Yahyā b. Ibrāhim al-Ḥakīm, but the official catalogue gives no author’s name, and I think that Schroeder has in transcribing his notes confused this MS. with MS. Abd-al-Ḥamīd, No. 561 (see below). The opening leaves of the MS. are lost, and the MS. itself as a result throws no light upon the question of authorship. It was copied 21 Ramadān 618/1221 by Muḥammad b. Hawā b. ‘Othmān, the mueddib, as appears from the conclusion on f. 77 b.

In addition to the loss of leaves at the commencement of the MS., there is a gap between ff. 14 and 15 (f. 14 ends with the chapter-heading, ‘Chapter of the’ibdīyāt which the different classes of chess-players have chosen,’ and 15 begins in the middle of a problem solution). The text of this MS. is in the main identical with that of AH, without retaining the order of that MS., and the seven pages of poems (ff. 60–62 a) all occur in AH.⁶

7. H = MS. John Rylands Library, Manchester, Arab. 59.

8. Z = MS. Abd-al-Ḥamīd I, Constantinople, No. 561.

These are two MSS. of the same Arabic work, the Nuzhat al-arbāb al-‘aqūl fī’sh-shaṭranj al-manqūl (‘The delight of the intelligent, a description of chess’), by Abū Zakarīyā Yaḥyā b. Ibrāhīm al-Hakīm. The author flourished in the middle of the fourteenth century. He quotes from the great dictionary of his contemporary al-Fīrūzābādī (I). 817/1414, aged 85), the al-Qāmūs (H, f. 4 a), and there is a quotation from al-Ḥakīm’s book in b. Abī Ḥajala’s work, which will be described next. Neither MS. is dated, but H is ascribed to the latter half of the fifteenth century. Z is a modern MS., written perhaps towards the end of the eighteenth century.

H consists of 57 paper leaves, 175 mm. by 130. This MS. and the companion chess MS. in the Rylands Library (Man., see below) were brought to England from Damascus in the eighteenth century, and formed part of the collection of J. G. Richards, until in 1806 they passed into the possession of John Fiott, of St. John’s College, Cambridge, who subsequently took the name of Lee on inheriting property from his mother’s family.⁷ Nathaniel Bland borrowed them from Dr. Lee for use in the preparation of his paper on Persian Chess (London, 1850), but failed to return them, and subsequently efforts to recover them which were made at the instance of Prof. Duncan Forbes between 1855 and 1860 proved fruitless. Bland’s Oriental library was sold en bloc in 1866 to the Earl of Crawford and Balcarres, and the ‘Lee MSS.’ passed into the Haigh Hall Library, and were duly entered in the printed Hand-list to the Oriental MSS. of that library. In 1906 Lord Crawford’s Oriental MSS. were purchased by Mrs. Rylands, who subsequently placed them in the noble library which she had founded in memory of her husband.

Z is a paper MS. of 56 leaves, 143 mm. by 70. It has no title, and the entry in the official catalogue (‘561. Risāla fi‘sh-shaṭranj, one volume in Arabic, by Yaḥyā b. Muḥammad aṣ-Ṣūlī’) is unwarranted by anything in the MS., which names al-Ḥakīm as its author in the opening sentence.

Al-Ḥakīm’s work is based upon the works of al-‘Adlī and aṣ-Ṣūlī, and carefully discriminates between the problem material which was taken from each of these lost works. The introduction contains a large number of stories relating to chess which are not given in any of the older MSS., and the conclusion contains a number of chess-poems, together with sections on the game at odds, and on the technical terms used in chess, and some notes on a group of famous players of the end of the 12th c.

The two MSS. are in the main identical in contents, with some variation in the order of the problems which is sufficient to show that Z is not a transcript from H. Z also omits one of the Knight’s tours included in H.

9. Man. = MS. John Rylands Library, Manchester, Arab. 93.

A MS. of 89 quarto leaves, 174 mm. by 130, copied 850/1446, bearing the title Kitāb ’anmūdhaj al-qitāl fī la‘b ash-shaṭranj (‘Book of the examples of warfare in the game of chess’), by Shihābaddīn Abū’l-‘Abbās Aḥmad b. Yaḥyā b. Abī Ḥajala at-Tilimsānī alḤ-anbalī (B. 725/1325, D. of the plague, 776/1375).

This work is written in eight chapters with introduction and conclusion. Each chapter concludes with five diagrams, (1) an Opening, (2) and (3) two won problems, (4) and (5) two drawn problems. The introduction deals with the stories of early Muslim players, the question whether chess was makrūh or ḥarām (see Ch. XI), under what conditions Muslims might play the game, and the correct spelling of the word shaṭranj. Ch. i (f. 14 b) treats of the invention of chess; ch. ii (f. 26 a) of the classes of players, the values of the pieces, and the symbolism of the game: ch. iii (f. 31 a) contains a long extract from aṣ-Ṣūlī giving maxims and advice for chess-players, to which b. Abi Ḥajala added a critical commentary. Aṣ-Ṣūlī’s advice is very similar to that contained in the treatise by al-Lajlaj which is contained in AH. Ch. v (f. 41 b) deals with the temperaments of chess-players: ch. vi (f. 46 b) contains quotations in praise and dispraise of chess, among others one on f. 47 b is said to be taken from the K. al-manṣūbāt of Abū Zakarīyā Yaḥyā b. Ibrāhīm al-Ḥakīm, the author of the MSS. H and Z. Ch. vii (f. 54 a) treats of the varieties of chess and exercises or puzzles (see Ch. XV); ch. viii (f. 76 b) is a poetical miscellany of extracts relating to chess; and the conclusion (f. 81 a) is a maqāma shaṭranjīya, a prose essay in the elaborate style set by al-Ḥarīrī (D. 515 or 516/1122), and dedicated to the Sultan al-Malik aṣ-Ṣāliḥ of Mārdīn.

One of the most valuable features in this MS. is the information which it supplies as to the nature of the traditional diagrams of normal positions in the Openings.

10. A1. = The chess chapters in al-Amulī’s encyclopaedia.

The encyclopaedic Persian Nafā’is al-funūn fī ‘arā’is al-‘uyūn (‘Treasury of the Sciences’) of Muḥammad b. Maḥmūd al-Amulī (D. 753/1352) concludes with three chapters on chess. MSS. of this work are common in European libraries, though the chess chapters, as the last in the work, are often copied perfunctorily, and, if the MS. be defective or unfinished, they generally suffer. I have used eight MSS., four in the Bodleian, two in the British Museum, one in the India Office Library, and one in the Imperial Library, Vienna. None gives the diagrams complete.

The first chapter is introductory, dealing with the Indian invention of the game; the second chapter deals with the derived games of chess; the third with problems; and the work concludes with ‘some amusing and sensible remarks respecting the morals and social observances or amenities of the Royal Game’.

11. RAS = MS. Royal Asiatic Society, Persian, No. 211.

A MS. of 64 quarto leaves, 9 in. by 7 in., written in nashkī hand. The MS. is imperfect at the end, and the leaves are in some confusion. It was presented to the Society by David Price, and was formerly catalogued as No. 260.

This appears to be Ḥājjī Khalīfa’s third chess work; since it is mainly devoted to the praise of ‘Alī ash-Shaṭranjī, the great player at Tīmūr’s court, it has been suggested that this player may be the author of the MS., in which case it may be the work mentioned by b. ‘Arabshāh.

Ff. 1 b–32 a are occupied by diagrams, one a page, with actual players depicted to the right and left of the board, which is placed with the files vertical (in my extracts from this MS. I regard the h-line as being at the foot of the page). The whole is illuminated, but the pieces are merely indicated by their names in red and black ink. At the head of the page is the heading of the problem, with the name of the player to whom the author has ascribed it. This MS. differs from all other older Muslim MSS. in giving no solutions to the problems.

The remainder of the MS., according to Forbes, can be rearranged to give (1) a single leaf forming a portion of the preface, in which the writer boastfully records his own achievements at chess, (2) 12 leaves on the beneficial effects of chess, (3) 7 leaves with a diagram on the Complete chess (Tīmūr’s chess), for which see Chapter XVI below, (4) 7 leaves on the invention of the ordinary chess in India, (5) 3 leaves containing sections on the relative values of the chess-pieces, on the gradation of odds, and on End-game decisions. See Bland (1–17) for a fuller account of the MS.

The MS. is probably of the 16th century.

12. F = MS. Nūri Osmānīye, Stambul, No. 4073.

13. Q = MS. Munich, 250. 25 Quatr.

These are two MSS. of the Shaṭranj nāma-i kabīr of the noted Turkish poet Firdawsī at-Tahīhal, the author of the immense Sulaiman nāma, a poem which, according to the present work (F, f. 7 b), filled 366 volumes, and contained 1,838 chapters and 890,000 verses. The chess work was compiled at Balakasri in Liva Karasi for the Sultan Bāyazīd II (A.D. 1481–1512), after the completion of the vast epic.

F is a MS. of 94 leaves, which was discovered by Dr. Schroeder, and is No. XXII of v. d. Linde’s list (Qst., 398 seq.). It was completed 907/1503 (f. 94 a). Q, a MS. of 87 leaves, 251 mm. by 180, also belongs to the sixteenth century, and was in Egypt from 1553 until the Napoleonic invasion. This MS. has several leaves missing. There are gaps between ff. 29 and 30, 41 and 42, 60 and 61, 62 and 63, 69 and 70, 73 and 74, 77 and 78, and the concluding leaves are missing.

Firdawsī arranged his work in eight chapters, in agreement with the eight squares on the edge of the chessboard. To these must be added a lengthy introduction treating of the history of the composition of the book, and a shorter conclusion. Chapter i treats of the invention of chess and legends associating the prophet Idrīs, Jimjīd, and Solomon with chess; chapter ii deals with the mastership of Lajlāj, later named in full as Abū’l-Faraj b. al-Muẓaffar b. Sa‘īd; chapter iii treats of the match which Lajlāj played with Buzūrjmihr in the presence of Nūshīrwān; chapter iv gives the rules and maxims as laid down by the prophet Idrīs; chapter v tells the story of the tribute of the grains of corn which Lajlāj demanded from Nūshīrwān, and adds chess legends of Iskander (Alexander the Great) and other rulers; chapter vi gives the ta‘bīyāt, and chapter vii the manṣūbāt; chapter viii discusses the legality of chess-playing. Almost every chapter concludes with a poem, and every problem with a couplet.

Firdawsī’s work is in the main a compilation from other works. He specially notes (F, f. 11a) his indebtedness to the Shāhnāma of his great namesake, to the ‘Ajā’ib makhlūqāt (probably by aṣ-Ṣafadī, 896/1490), to the Qābūs-nāma (written A.D. 1082–3 by ‘Unsuru’l-Ma‘ālī Kaykā’ūs, Prince of Ṭabaristān), to the Gharā’ib mawjūdāt, and to the Ikhwān aṣ-ṣafā.

14. R = MS. Rustem Pasha, Constantinople, No. 375.

A paper MS. of 90 leaves, 21 cm. by 15.2, which forms part of a MS. of miscellaneous contents which was written by Aḥmad b. Aḥmad al-Muhtār al-Ḥanafī al-Miṣrī at Balat, Stambul, in 983/1575. It contains nothing but problems, one to the page, with solutions.

15. S = MS. Bodleian Iib., Oxford, Arab. Pocock 16.

A small parchment MS., completed 979/1579, containing three treatises by Muḥammad b. Aḥmad b. Muḥammad b. Jamāladdīn b. Sukaikir ad-Dimashqī, preacher of the mosque al-‘Ādilīya at Halab (Aleppo, in Syria) (D. 987/1579). The first two treatises are theological; the third, which extends from f. 22 a to f. 39 b, treats of chess under the title Nafḥāt kimā’īm al-ward fī tafḍīl ash-shaṭranj ‘alā’n-nard (‘The fragrance of the rose: on the superiority of chess over nard’). The MS. was once in the possession of Dr. Hyde, who made large use of it for his Mandragorias.

The MS. discusses the lawfulness of chess-playing, summarizing for the purpose aṣ-Ṣūlī’s collection of traditions, but while giving the usual legends as to the invention of chess—in his day there were people who thought that aṣ-Ṣūlī had invented chess—b. Sukaikir adds some interesting particulars as to notable feats at chess, some of which had taken place in his presence. He only gives 10 problems, omitting the solutions, but indicating the number of moves to be taken. He gives a number of impromptu verses on chess.

16. Y = MS. Brit. Mus., Add. 16856.

A Persian MS. of 62 leaves, 10 ins. by 6, written in a neat nestalik hand, with ‘Unwān and gold-ruled borders, dated 1021/1612, from the library of Col. Wm. Yule. It is a Persian translation by Muḥammad b. Ḥusām ad-Daula of the Arabic work K. al-munjiḥ fī ‘ilm ash-shaṭranj (‘A book to lead to success in the knowledge of chess’) by Muḥammad b. ‘Omar Kajīnā, a work stated in the preface to be the most useful treatise on chess. As there was, however, only one copy of it in the land, and that an incorrect one, it appeared desirable to make an abridged version of it in Persian, and the author performed that task at the order of a sovereign whose titles and epithets are given at length, but whose proper name does not appear.⁸

It is divided into fourteen chapters, but the copy, although showing no sign of loss since it left the writer’s hand, is not complete; only three lines of chapter ix are given, and the termination of chapter xi and the whole of chapters xii and xiii are missing.

The chapters deal, i, with stories of early Muslim players; ii, with the question of the lawfulness of chess-playing; iii, with the advantages of chess; iv, with the invention of chess; v, with technicalities of the game; vi, with the etiquette of play; vii, with maxims for players; viii, with the Endgame decisions; ix, with the ta‘bīyāt; x, with conditional problems; xi, with problems in general; and xiv, with blindfold chess.

With this work I complete the list of the Oriental MSS. of the older chess which I have made the basis of my chapters on the practical game and the Muslim manṣūbāt. In the case of BM, H, Man., RAS, S, and Y, I have been able to refer to the original MSS. themselves. For the opportunity of consulting the other MSS. I am indebted to the generosity of Mr. J. G. White. He has placed at my service his photographic copies of AH, V, Z, RAS, Q, and R, and modern transcripts of AH, C, BM, AE, V, and F.

I now give a tabular summary of the contents of these MSS.

CONTENTS OF THE MSS. DESCRIBED ABOVE.

Table of Contents

Its Persian ancestry.—The date of introduction.—The legal status of chess.—Early Muhammadan chess-players.—The game during the Umayyad and ‘Abbāsid caliphates.—Aṣ-Ṣūlī.—Later references.—Aṣ-Ṣafadī.—Chess at the court of Tīmūr.—Chess in Damascus in the sixteenth century.

That Islam derived its knowledge of chess from Persia cannot be disputed for a moment. The Arabic historians who make any reference to the matter, however much they may differ as to the ultimate origin of the game, agree in stating categorically and as an undisputed fact, ‘We learnt chess from the Persians.’ Of greater weight is the philological evidence derived from the Arabic nomenclature of chess. The Persian consonant ch has never existed in Arabic, and had to be represented in Arabic by sh or ṣ. Examples of both will be found below, p. 217, n. 20. The Arabic letter j (= Hebrew gimel), which perhaps still retained the original sound of the ‘hard’ g under the early caliphate, was used to represent the ‘hard’ Persian g. The Arabic j is still pronounced as ‘hard’ g in Egypt; elsewhere it is pronounced as the English (or even French) j. Shaṭranj,¹ the Arabic name of chess, is accordingly the regular Arabicized form of the Persian chatrang. With one exception, the Persian names of the chess-pieces are retained in Arabic, and shāh, firzān, fīl, rukhkh, and baidaq or baidhaq (pron. baizaq) are the regular Arabicized forms of the Persian shāh, farzīn, pīl, rukh, and payādah. The ‘horse’ alone received a native name, t e Persian asp being translated by the Arabic faras.

Nor can there be much doubt that the introduction of chess was a result of the conquest of Persia which took place between the years A.D. 638 and 651, in the caliphate of ‘Omar b. al-Khaṭṭāb, the second of the four orthodox caliphs, and thus some years after the death of Muḥammad himself (A.D. 632). Most probably the prophet had never heard of the existence of chess, since the Muhammadan jurists have been unable to settle the question of the legality of chess-playing by any direct decision of Muḥammad as recorded in the Qar’ān, or in authentic tradition. Such at any rate was the opinion of the earlier lawyers.

This question of the legal position of chess-playing exercised the early Muslim lawyers not a little. The whole possibility of a Muslim chess depended upon the decision that was reached. Muslim law is far wider in scope than anything that the Western world has ever known. As D. B. Macdonald puts it in his Development of Muslim Theology, Jurisprudence, and Constitutional Theory (London, 1903, p. 66), ‘Muslim law in the most absolute sense fits the old definition, and is the science of all things, human and divine. It tells us what we must render to Caesar and what to God, what to ourselves, and what to our fellows. The bounds of the Platonic definition of rendering to each man his due it utterly shatters. While Muslim theology defines everything that a man shall believe of things in heaven and in earth and beneath the earth—and this is no flat rhetoric—Muslim law prescribes everything that a man shall do to God, to his neighbour, and to himself. It takes all duty for its portion and defines all actions in terms of duty.’ Nor was this any empty claim. A Muslim’s citizenship depends upon his character, as judged by his conformity to the letter of the law, and it is only the evidence of a man of ‘blameless life’ that possesses any validity in a court of law. If the practice of chess was established to be illegal, no true Muslim could be a chess-player. It became, therefore, a matter of importance to ascertain the legal position of chess and chess-playing.

It was not, however, until the second century of Islam that any serious attempt was made to systematize and codify Muslim law. Prior to this lawyers had been mainly opportunists, though the seeds of the broad separation of Muhammadans into Sunnites—those who accepted the caliphate de facto—and Shi‘ites—those who upheld the right of the descendants of ‘Alī and his wife Fāṭima, the Prophet’s daughter, were already there. But in the second and third centuries Sunnite law was systematized by a number of schools or sects, of which four stand out above the others. These are the Ḥanīfite, the Mālikite, the Shāfi‘ite, and the Ḥanbalite schools, so called from their respective founders, Abū Ḥanīfa (D. 150/767), Malik b. Anas (D. 179/795), ash-Shāfi‘ī (D. 204/820), and Aḥmad b. Ḥanbal (D. 241/855). To one or other of these sects practically every Sunnite Muslim belongs to-day, and in broad outline Sin‘ite law is not very dissimilar.

Muslim law divides all actions into five classes—(1) necessary actions (farḍ, wājib), the omission of which is punished, and the performance of which is rewarded; (2) recommended actions (mandūb, mustaḥabb), the omission of which entails no penalty, but the performance of which is rewarded; (3) permitted actions (jā’lz, mubaḥ) which are indifferent legally; (4) disliked actions (makrūh) which are disapproved but not under penalty; and (5) forbidden actions (ḥarām), the performance of which is punished by law.² The criteria for the proper classification of actions have varied somewhat from time to time, and with different schools, but all agree that the final criterion is the Qur’ān, and that next in importance comes the evidence of a clear and authentic tradition of Muḥammad or of the earliest age.³

Chess is mentioned nowhere by name in the Qur’ān, but, adopting the principle of analogy (qiyās) by which the doubt could be resolved by a decision on some similar case, appeal was made to Sura V. 92, a chapter that belongs to the Medina or last period of Muḥammad’s life. In this verse we read—

O true believers, surely wine and lots (maisīr) and images (‘anṣāb) and divining-arrows (’azlām)⁴ are an abomination of the works of Satan, therefore avoid ye them that ye may prosper.

It is by extending the condemnation of lots—maisīr—and images ‘anṣāb—that the attempt has been made to condemn chess and chess-playing. There is fair agreement among the commentators that maisīr was intended to include every game which is subject to hazard or chance, or which is played for money or a stake. It is on this verse that the prohibition of nard (tables, backgammon),⁵ and the later-discovered games of cards is based. There is, however, a tradition which is preserved by al-Baihaqī (D. 458/1066) that the caliph ‘Alī once described chess as the Persian maisīr, though the genuineness of the tradition is disputed by other writers—b. Sukaikir, for instance. The noted Ḥanbalite b. Taimīya (D. 728/1328) makes the sensible distinction that in chess it is only the playing for money that is maisīr, and quotes the opinion of Mālik b. Anas that the stake made chess a far worse game than nard. The Sunnite Muslim sees a prohibition of carved chess-pieces which actually reproduce the King, Elephants, Horses, &c., in the prohibition of images.⁶ Persian commentators, however, have explained the term as referring to idols,⁷ and the Shi‘ite and Moghul chess-players have no objection to using real carved chessmen. The Sunnite player, on the contrary, will only use pieces of a conventional type in which it is impossible to see any resemblance, to any living creature.

In the second place, the lawyers turned to the traditions (ḥadih) of Muḥammad and his immediate companions, in order to deduce what their practice in the matter of games was. And here at the outset lay the difficulty of settling the genuineness or otherwise of the tradition. Islam was flooded with traditions by the end of the second century,⁸ and the vast majority of these were forgeries. Only the crudest tests could be applied in an age that had no appreciation of the science of historical criticism. But crude as the tests were, they disposed of ninety-nine per cent of the traditions.⁹ And in the winnowed material three traditions survived which dealt with Muḥammad’s attitude towards recreations. One of these emphasizes his hatred of games of chance, another shows his approval of martial exercises with lance or bow, and the third preserves a statement that a believer should restrict his amusements to his horse, his bow, and his wife or wives.

These traditions form the basis of the discussion as to the status of chess in the works of the founders of the four great schools. Abū Ḥanīfa reduces the question to a dilemma: either the game is played for a stake, or for amusement. In the first case it is forbidden by the Qur’ān, in the second it is not one of the three forms of recreation allowed by Muḥammad. Chess, nard, and fourteen¹⁰ are all clearly illegal. There is, however, a difference of degree. Chess is only disapproved (makrūh), not forbidden (ḥarām), as is nard. It is a sin that leads into error, and Abū Ḥanīfa did not himself refuse to greet a chess-player when at his game. The Ḥanīfite code was the official ‘Abbāsid canon, but later lawyers had to exercise considerable casuistry to reconcile their law with the wishes of the caliphs.¹¹

Mālik b. Anas and b. Ḥanbal took a more hostile view. In the K. al-muwaṭṭa, of Mālik’s Spanish pupil Yaḥyā b. Yaḥyā (D. 234/848) there is added to the citation of the tradition in which Muḥammad interdicts games of chance the following reminiscence of his master’s hatred of chess:

I heard Mālik say that there was nothing good about chess. He pronounced it ḥarām. I heard him denounce chess-playing and other vanities as ḥarām, quoting Sura X. 33, ‘When the truth has been scorned, what is left except error.’

Indeed Mālik held that chess was far worse than nard, since the game exercised a far greater fascination over its players. The Ḥanbalite school were equally opposed to chess, but they took the more natural view that nard was still worse.

Ash-Shāfi‘ī enunciated a more liberal view. He found Abū Ḥanīfa’s dilemma defective, since he claimed that chess is an image of war, and it is possible to play chess not for a stake, not for pure recreation, but as a mental exercise for the solution of military tactics. When played for this last purpose, he denied that the player was doing anything illegal. According to al-Māwardī (D. 450/1058) he regarded chess as makrūh, not because it leads into error—that ash-Shāfi‘ī denied—but as a sin of recreation. And provided the player took care that his fondness for chess did not cause him to break any other rule of life, he saw no harm in playing. Ash-Shāfi‘ī, indeed, played chess himself, defending his practice by the example of many of the companions and tābi‘s. The chess-players naturally attached great importance to the example of these early players, to whom all the legal schools looked back with reverence, and all the MSS. contain in more or less detail the traditions that enshrine the record of this or that tābi‘s approval or practice of chess. The great master and historian aṣ-Ṣūlī gave these traditions in text (matn) and chain of authority (‘isnād), and the MSS. AH and C have preserved his work for us. I shall make use of his traditions in this chapter. They contain the germs of the conditions which ash-Shāfi‘ī finally laid down as defining the lawfulness of play. These were four in number, the game must not be played for a stake, and no money must be paid in connexion with the game, the game must in no way be allowed to interfere with the regular performance of prayer or other religious duty, the player must refrain from angry and improper language, and the game must not be played in the street or other public place. It is obvious that these conditions are not compelled by any inherent quality in chess, but are due to the weakness and depravity of human nature. This is b. Sukaikir’s contention, that there is nothing wrong in the game itself, but only in the circumstances of play. He claims, therefore, that common-sense ought to justify the game, while he reluctantly admits that the general consensus of legal opinion is hostile.¹²

If we omit a very doubtful tradition that ascribes the story given below in connexion with the caliph ‘Alī to the first caliph Abū-Bakr (D. 13/634), the first traditions that connect a caliph with the game relate to ‘Omar b. al-Khaṭṭāb, the father-in-law of Muḥammad (D. 23/643). A widely recorded tradition tells how he was once asked as to the legal status of chess. ‘What is chess?’ asked the caliph. He was told that there was once a queen whose son was slain in battle. His comrades hesitated to tell her the news, and when she asked how the battle had gone, they invented chess and showed it to her. By means of the explanation they conveyed the news of the prince’s death.¹³ ‘Omar listened to the tale, and then replied: ‘There is nothing wrong in it; it has to do with war.’ The fact that ‘Omar once greeted Hilāl b. Khasīb, a maula (dependent, client) of Sulaiman b. Yasbār (D. 107/725), the great tābi‘of Medina,¹⁴ while he was engaged in chess is handed down in a tradition with particularly good’isnad.¹⁵

The caliph ‘Alī b. Abū Tālib (D. 40/660), the son-in-law of Muḥammad, is connected with the following story, the genuine nature of which was allowed by the traditionists:

‘Alī once chanced to pass by some people who were playing at chess, and asked them, ‘What images are these upon which you are gazing so intently?’, for they were quite new to him, having only lately been introduced from Persia, and the Pawns were soldiers, and the Elephants and Horses were so depicted according to the custom of the Persians.

It is inferred from this that ‘Alī only objected to the carved chessmen and not to the game itself, and it is in deference to this that the Sunnite Muslims use men of a conventional pattern.

Al-Māwardī (D. 450/1058) quotes traditions that connect several of the ‘Companions’ (aṣḥāb) with chess. Abū Huraira (D. 57/676–7),¹⁶ ‘Abdallāh b. ‘Abbās, and ‘Abdallāh b. Zubair are stated to have been seen to play chess, while al-Ḥusain (sl. 68/610), the ill-fated son of the caliph ‘Alī, is recorded to have played with his children, and also to have watched a game and to have prompted the players.

The traditions regarding the tābi‘s are equally trivial in detail, and their main interest consists in the evidence they afford for the practice of chess in the first centuries of Islam. Since some of these early players are said to have played the game blindfold, it is reasonable to conclude that the standard of play must have been fairly high. The cosmopolitan nature of Islam is well illustrated by the nationalities of these chess-players.

The list includes the names of Sa‘īd b. al-Musayyib¹⁷ (D. 91/709–10), of Medina, an Arab, who played in public and declared the game permissible provided there was no stake; ‘Alī b. al-Ḥusain Zain al-‘ābidīn (D. 94/712–3), one of the Imams of the Shi‘ites, whose father was, as already mentioned, a chess-player, and whose mother according to legend was Shahr-bānū, the daughter of Yazdigird III, the last of the Sāsānian kings of Persia; Sa‘īd b. Jubair (ex. 95/714), a negro, who excelled in blindfold play; Ibrāhīm b. Ṭalha b. ‘Obaidallāh (D. 98/717), the son of one of Muḥammad’s earliest converts, who had been seen to play chess in public in Medina; al-Qāsim b. Muḥammad¹⁸ (D. 101/719–20), by his father a grandson of the caliph Abū-Bakr, and by his mother of Yazdigird, who once rebuked some chessplayers for using figures (sūrun) for pieces (dawābb); ash-Sha‘bī (D. 108/722–3), of Persian descent, who played chess and nard for a stake and forgot the hour of prayer, and played in the street, covering his head so that he should not be known; ‘Ikrima (D. 107/725–6), a Berber; Muḥammad b. Sīrīn (D. 110/728–29), a Persian, who was famous for his interpretation of dreams, and could also play chess blindfold; al-Ḥasan al-Baṣrī (B. 110/728), who saw no harm in chess provided there was no stake and no neglect of the times of prayer; ‘Aṭā‘a (D. 115/733–4), a deformed mulatto; az-Zuhrī (D. 124/742), the great lawyer of the Umayyad period; Muḥammad al-Munkadir (D. 131/748–9); Rabī‘a ar-Rai (D. 136/753–4), of Persian descent; Hishām b. ‘Urwa (D. 146/763–4), another blindfold player, whose three granddaughters Safī‘a, A‘īsha, and ‘Ubaida also all appear as chess-players; al-A‘amash (D. 148/765), a Persian; and Abū ‘Aun (D. 151/768), another Persian. Although Mālik b. Anas was so opposed to chess, he numbered among his friends a chess-player, al-Mughīra b. ‘Abdarraḥmān, and his own son Yaḥyā b. Mālik b. Anas, who was a lawyer in Medina, played chess in his home. Finally, the great lawyer, ash-Shāfi‘ī, is credited with skill in blindfold play.

Probably the most interesting in these names is that of Sa‘īd b. Jubair. According to b. Taimīya (Man., f. 10 b), he gave the following curious reason for his playing chess. He had reason to believe that al-Ḥajjāj desired to appoint him qāḍī, and, fearing that the patronage of this noted man would be detrimental to his piety, he took up chess in order to disqualify himself for the post. Chess-playing he regarded as the less of the two evils, and since acts are to be judged by the intention, even a more heinous sin would have been permissible in his necessity. He was only forty-nine when the same al-Ḥajjāj put him to death for taking part in a revolt against ‘Abdalmalik b. Marwān. His murderer is said to have dreamt that God would kill him once for every man he had killed in his ruthless career, but seventy times for the death of Sa‘īd b. Jubair. Other traditions in AH tell us that Sa‘īd had played chess all his life, that he played with equal ease whether he saw the board or not, and that his method of playing blindfold was to turn his back on the board; then he would ask the slave who attended him what his opponent had moved, next he bade the slave ‘move such and such a man’. His name is the earliest one that is associated with play without the use of a material board, but he may have had many followers among the Muslim players. Other references to players who could play blindfold are given later in this chapter, and a Muslim whose name is given as Buzecca or Borzaga is mentioned as the first exponent of the art of blindfold play in Europe. This player visited Florence in 1265.¹⁹

The earliest of the Umayyad caliphs who is associated with chess is ‘Abdalmalik b. Marwān (D. 86/705). An earlier caliph, Yazīd I b. Mu‘āwiya (D. 64/683), the hated murderer of the Imām Ḥusain b. ‘Alī, is stated by b. Khallikān to have been a nard-player, and accordingly a man whom it was legally permissible to curse. There are three stories of ‘Abdalmalik in H (ff. 8 a, 11 a, and 14 b). They merely exhibit the caliph as a chess-player, but one brings in the noted poet al-Akhṭal (D. 92/710), and another tells how ash-Sha‘bī, whom we have already heard of as an inveterate chess-player, once asked the caliph if he was not ashamed of playing. The caliph answered by some questions. Was the game ḥarām? or maisīr or ’ansāb? Since ash-Sha‘bī could only answer all these in the negative, the caliph continued to play. AH, f. 12 b, has a curious story, which the later K. al-‘uyūn, a generally trustworthy history of the 5th or 6th c. of Islam (say A.D. 1150–1250, ed. de Goeje in Fragmenta Hist. Arab., Lugd. Bat., 1871, p. 102), repeats in connexion with ‘Abdalmalik’s younger son, the caliph Hishām (D. 125/742), while MS. Brit. Mus. Add. 7320, f. 42 b, which has been identified by H. F. Amedroz as the work of b. al-Jauzī (‘An unidentified MS. by ibn al-Jauzī’, JRAS., Jan. 1907, see p. 865), attributes it to the elder son, the caliph Walīd I b. ‘Abdalmalik (D. 96/714). The caliph was once engaged in playing chess when a visitor,²⁰ a Syrian, was announced. The caliph ordered a slave to cover over the chessboard, and the visitor was allowed to enter. The caliph then proceeded to examine his guest in order to find out how far he was instructed in the Muslim religion, and, discovering that he was quite unlearned, he bade the slave uncover the board, and resumed his game, for ‘there is nothing forbidden to the uneducated’. This story is gravely told by aṣ-Ṣūlī as evidence for the legality of chess-playing. Its unsatisfactory nature and the fact that it is cited and not suppressed in the chess MSS. is in favour of its genuineness.

The chess MS. Y and some later (for the more part Indian) chess works give a story of ‘Abdalmalik’s son and successor Walīd I (D. 96/714). He was once playing chess with a courtier who purposely played negligently to avoid beating the caliph. On discovering this the latter took umbrage, and broke his flatterer’s head with a blow with his firzān, saying, ‘Woe be to you! are you playing chess, and in your senses?’ The silence of earlier works tells against this story.

A thoroughly satisfactory reference of about this time is to be found in a passage in one of the poems of the noted poet al-Farazdaq (D. c. 110/728). This is the more important since there is an allusion to a technicality of chess which would not have been appropriate unless the game were fairly generally known. It must take time for a peculiarity of a game to become sufficiently known to take its place in literary idiom. The couplet in question runs:

And, as for us, if Tamīn reckons his ancestors in the rank of the forelocks of the noblest victors of the race-course, I keep you from your inheritance and from the royal crown so that, hindered by my arm, you remain a Pawn (baidaq) among the Pawns (bayādiq).

—an allusion to the promotion of the Pawn when it reaches the end of the board.²¹ So it is interpreted by al-Jawālīqī (D. 539/1145) in K. al-mu‘arrab, a work on Arabic loan-words which has been edited by Sachau, Leipzig, 1867, where the verse is quoted. Al-Jawālīqī states, rather loosely, that the Pawn which advances to the limit of the board ‘obtains the weapons of the King’.

Another contemporary poet, al-Aḥwaṣ (D. 110/728), is connected with chess in a passage in the K. al-aghānī of Abū‘l-Faraj (compiled A.D. 918–67) (ed. Bulaq, 1285, iv. 51). A certain ‘Abdalḥakam b. ‘Amr b. ‘Abdallāh b. Ṣafwān al-Hujamī possessed a house in Mecca where he kept sets of chess, nard, and merels,²² and books on all the sciences. The walls were provided with pegs, so that every one who entered could hang up his cloak. He was then expected to take a book, or to choose a game and to play with some other guest. Once ‘Abdalḥakam came across a stranger in the Ka‘ba to whom he took a fancy. He brought him home with him, and after hanging up their cloaks he took down the chess and challenged him to a game. Just then the singer al-Abjar entered, and greeted the unknown with, ‘Hullo, heretic!’ and to ‘Abdalḥakam’s astonishment presented him as the Medinese poet al-Aḥwaṣ. This incident must have taken place after al-Aḥwaṣ’s return from banishment in 101/719.

Ar-Rāghib (D. 502/1108) in his K. muḥāḍarāt al-udabā’²³ relates that the Persian Abu Muslim (D. 137/754–5) once quoted a verse of one of the older poets in a new sense when he was checkmated in a game of chess.

We may safely assert that chess had already become a popular game throughout Islam, from Spain to the banks of the Indus, before the commencement of the ‘Abbāsid caliphate.

The only chess story that brings in the name of the second caliph of the new dynasty, al-Manṣūr (D. 158/775), that I have come across, occurs in the chess MS. H (f. 10 b). The vizier of this caliph, Abū Ayyūb al-Muriyānī (D. 154/771), had a friend who was a skilled chess-player. The MS. quotes a witty couplet which the latter wrote to the vizier, inviting him to a game of chess.

Al-Mahdī (D. 169/785), the third of the ‘Abbāsid caliphs, the son of al-Manṣūr and father of Hārūn ar-Rashīd, looked—at least officially—with disfavour upon chess. A letter of his, written in 169/780 to the people of Mecca, is given in Arabic text in Wüstenfeld’s Die Chroniken der Stadt Mekka (Leipzig, 1861 ff., iv. 168). In this the following passage occurs:

Facts about you have been reported to the Commander of the Faithful which he has heard with regret and which he condemns and abominates. He desires you to abandon these things, and directs you to do away with them, and to cleanse the Sanctuary of God from them. To these things belong …. the assembly of fools for nard, dicing, archery, chess, and all vanities that lead astray and from the remembrance of God, which interfere with the fulfilment of your duty to Him, and the performance of prayers in His mosques.

Notwithstanding this, chess must have been played at al-Mahdī’s court, for we know from the K. al-aghānī (ed. cit., xix. 69) that the poet Abū Ḥafṣ ‘Omar b. ‘Abdal‘azīz, of Persian ancestry, was educated there, and that he obtained his surname of ash-Shaṭraujī, the chess-player, from his fondness for and skill in chess. After al-Mahdī’s death he, remained in the service of the caliph’s daughter ‘Ulayya, who is remembered for her love of music. Abū Ḥafṣ also played chess blindfold.²⁴

Although the MS. V (f. 24 a) attributes a problem (No. 181 below) to al-Mahdī, with the unusual information that the position was not derived from an actual game, it does not follow that the ascription has any historical weight. The MSS. show an ever-growing tendency to assign the authorship of approved problems to noted characters, and their statements need to be treated with much caution.²⁵ In the present instance the ascription is in conflict with the evidence of the historian Muḥammad b. ‘Alī al-Miṣrī, as recorded in al-Maṣ’ūdī’s Murūj adh-dhahab (ed. cit., viii. 295):

Ar-Rashīd was the first caliph to establish the game of assauljān (a Per. ball-game like polo) in the field, the use of the bow, and practice with the lance, with the ball, and rackets; he recompensed those who distinguished themselves in the different exercises, and people followed his example. He was also the first of the ‘Abbāsid caliphs to play chess and nard. He favoured good players and granted them pensions.

I have already quoted the letters that passed between Hārūn ar-Rashīd (170/786–193/809) and Nicephorus in 802. This is the only allusion to chess in Arabic historical works in which Hārūn is concerned. The occasional chess passages in that well-known compilation from early and late sources, the Alf laila walaila, ‘the Thousand and one Nights,’²⁶ are naturally of an unhistorical character, and can only be accepted for the Mamlūk period during which the collection of tales took its present shape in Egypt. The chess MS. H is the only one of those that I have used which contains much to connect Hārūn with chess, and none of its seven stories²⁷ has any real importance, apart from the impromptu verses to which they gave occasion. Four stories show the caliph in an inquiring mood. He asked his physician, b. Māsawaihi (D. 243/857), whether chess could be played during illness, and received the answer that it was generally suitable, but that at certain times—all detailed—it was inadvisable to play. Another time, on a wet day, he asked Yaḥyā b. Aktham the qāḍī (D. 242/847) what could be done on such a day, and received an enigmatical reply, which was interpreted as meaning to drink wine and play chess. On a third occasion he asked b. Māsawaihi what he thought of chess, and was told it was legally permissible; and on a fourth he started a controversy between the great Ḥanīfite, Abū Yūsuf the qāḍī (D. 182/795), and a Mālikite, Yaḥyā b. Bakair, on the same point. At first Abū Yūsuf defended the legality of chess, but when Yaḥyā declared that he had heard Mālik b. Anas forbid chess and reject the evidence of chessplayers, he gave up his contention, and agreed that Mālik’s opinion settled the matter. Another story tells the history of a slave girl who was famed for her skill at chess. Hārūn bought her for 10,000 dinars and proceeded to try conclusions with her at chess. He lost three games in succession, and when the slave was asked to choose her reward, she begged forgiveness for a certain Aḥmad b. al-Amīn. In these stories the noted poet Abū Nuwās appears as an intimate friend of the caliph. Another of Hārūn’s friends bore the name of Muḥammad al-Baid’aq, where the surname is derived from the name of the chess-pawn, and was given because the man was little of stature.²⁸

Hārūn’s eldest son and successor, al-Amīn (D. 198/813),²⁹ was also a chessplayer. Ar-Rāghib tells an amusing story of this caliph and the musician Isḥāq al-Mauṣilī (D. 235/849–50)³⁰ in the K. muḥāḍarāt al-udabā’, a work of which I have already made use. Al-Amîn and Isḥāq were once playing chess, and the latter had wagered his cloak on the game. The caliph won, but hesitated to take his opponent’s cloak, until the happy idea occurred to him to give up his own cloak as a gift. Al-Amīn’s fondness for chess led him to indulge in the game at unseasonable times. At the critical point of the siege of Baghdād, when the city was on the verge of capture, the messenger who was sent to the caliph to advise him of his peril found him deep in chess with his favourite Kauthar. ‘O Commander of the Faithful,’ he exclaimed, ‘this is not the time to play, pray arise and attend to matters of more serious moment.’ ‘Patience, my friend,’ coolly replied the caliph, ‘I see that in a few moves I shall give Kauthar checkmate.”³¹

Al-Ma’mūn (D. 218/833), who succeeded his brother al-Amīn in the caliphate, was equally addicted to chess, though apparently with less success. ‘Strange that I who rule the world from the Indus in the East to Andalūs in the West cannot manage 32 chessmen in a space of two cubits by two,’ is the remark that aṣ-Ṣafadī records of this caliph. Al-Yazīdī (D. 310/922) is quoted by b. Badrūn and as-Ṣuyūṭī as giving Ma’mūn’s opinion that chess was more than a game, and that to play it was an excellent training for the mind. The caliph tried to improve the game by introducing some novelties, which never took root. He also insisted on his opponent playing his best. Thus in the MS. Y we read—

Al-Ma’mūn was one day playing with a courtier who appeared to be moving negligently in order to allow the caliph to win the game. Al-Ma’mūn perceived it, and in great wrath upset the board, exclaiming, ‘You want to treat me as a child, and to practice on my understanding.’ He then addressed the onlookers: ‘Bear witness to the vow which I now make that I will never play chess with this person again.’

But if al-Ma’mūn himself was only a weak player, he yet liked to have strong players about him. On his expedition from Khurāsān to Baghdād in 204/819 he watched Rabrab,³² Jābīr al-Kūfī, and ‘Abdalghaffār al-Anṣārī play. The presence of the caliph manifestly embarrassed the players. ‘Chess and reverence,’ observed al-Ma’mūn, ‘don’t seem to agree. They ought to talk together just as they would do if they were by themselves.’ This incident is most interesting, for Jābīr and Rabrab are named in the chess MSS. as belonging to the highest class of players, that of the ‘aliyāt or grandees. These MSS. give some End-game positions that are drawn from actual games between Rabrab and Abū’n-Na‘ām, whose name follows that of Rabrab in the list of ‘alīyāt. The names are plainly in chronological order, and this age of al-Ma’mūn must have been a notable one in the history of Muslim chess, since it saw three grandees of chess living at one time.

The next caliph—still another son of Hārūn—al-Mu‘taṣim (D. 227/842), possesses a chess reputation that appears to have no real basis.³³ The only certain chess fact of al-Mu‘taṣim’s caliphate is the appreciation of the function of the fīl in chess which I quote in Ch. XIII, which was pronounced by his famous vizier Muḥammad b. az-Zayyāt (ex. 233/847).

During the rule of al-Wāthiq (D. 227/842) and al-Mutawakkil (D. 232/847)³⁴ the great master al-‘Adlī ranked alone in the highest class of players. It was only towards the end of his life that a rival appeared to dispute his position in the person of ar-Rāzī. The match took place in the presence of al-Mutawakkil, and, by defeating his older opponent, ar-Rāzī was successful in establishing his claim to be ranked among the ‘alīyāt. Both players were chess authors, but while we possess large portions of al-‘Adlī’s work in the various MSS., all that has survived of ar-Rāzī’s work is a few opinions on the End-game, a few aphorisms, and a couple of problems. Notwithstanding this neglect, aṣ-Ṣūlī considered that ar-Rāzī was the greatest of his predecessors. Of al-‘Adlī he had a poorer opinion, and much of his own chess work took the form of a criticism of al-‘Adlī’s book.

Aṭ-Ṭabarī (D. 310/923) in his K. akhbār ar-rusul wal-mulūk (ed. Goege, 1881, iii. 1671)³⁵ describes how the caliph al-Mu‘tazz (D. 255/869) received the news of his predecessor and rival al-Musta‘īn’s defeat and death in 252/866. The caliph was seated at chess when a messenger arrived bringing the head of al-Musta‘īn. Al-Mu‘tazz paid no attention to the news until he had finished his game.³⁶

An incident that al-Maṣ‘ūdī (op. cit., viii. 13) tells of Aḥmad b. Mudabbir, collector of taxes in Palestine under al-Muhtadī (255–6/869–70) shows that wealthy people kept good chess-players in their households. A certain b. Darraj intruded into b. Mudabbir’s house on one occasion and was discovered among the company. His host addressed him thus—

A parasite may be pardoned his intrusion upon other people’s society whereby he disturbs the charm of their intimacy and discovers their secrets, but only on the condition that he is endowed with certain talents, as a knowledge of chess or nard, or the ability to play the lute or guitar (tonbūr).

The stranger replied that he excelled in all these accomplishments, so b. Mudabbir ordered one of his pages to play the intruder at chess. The latter asked what reward he would get if he proved successful. He was promised 1,000 dirhems if he proved himself superior to all the company in his accomplishments. The money was brought and placed on the table, since the parasite said its presence would stimulate him to his best efforts. He won the game of chess and was about to take the money, when the doorkeeper, who saw a danger of punishment for his carelessness in allowing the stranger to enter, intervened, and said that he was sure another of the pages could beat the stranger at chess. This page was summoned and the stranger was beaten. He claimed a game at nard, first winning and then being beaten by a better player, and so the contest went on, the porter endeavouring to escape the consequences of his carelessness and the stranger to escape the thrashing he deserved for his impertinence.

Al-Mu‘taḍid, caliph 279/892–289/902, was also a chess-player. Al-Muṣ‘ūdī mentions (op. cit., viii. 271) that his vizier Qāsim b. ‘Obaidallāh once heard him quote a verse from b. Bassām during a game.

It was under the following caliph, al-Muktafī (289/902–295/908), that the historian Abū-Bakr Muḥammad b. Yaḥyā aṣ-Ṣūlī³⁷ first came into note as a chess-player of consummate skill. Ar-Rāzī was already dead, and no one had taken his place, when a certain al-Māwardī made his appearance at court and announced that his skill exceeded all that ar-Rāzī had ever possessed (H, f. 13 a). The caliph took al-Māwardī into favour, and when aṣ-Ṣūlī’s extraordinary talent at chess was reported to the caliph, he was not disposed to believe it. A match was arranged between the two players and took place in the caliph’s presence. Al-Muktafī was so led away by his partiality for his favourite that he openly encouraged him during the game. At first this embarrassed and confused aṣ-Ṣūlī, but he soon recovered his nerve, and finally defeated his adversary so completely that no one could doubt but that aṣ-Ṣūlī was by far the better player. When the caliph was thus convinced, he lost all his partiality for al-Māwardī, and said to him, ‘Your rose-water (māward) has turned to urine!’

The new grandee of chess was descended from Ṣul-takīn, a Turkish prince of Jurjān, whose ancestral home was situated at the south-east corner of the Caspian Sea, on the banks of the River Atrek. Yazīd b. Al-Muhallab converted the warrior during the conquest of Khurāsān. His grandson married a sister of the poet al-Ahnaf, and a son of this marriage, Ibrāhīm b. al-‘Abbās aṣ-Ṣūlī (D. 243/857), was known as a poet of some ability. Ibrāhīm’s nephew was the chess-player, who also proved himself a ready versifier and was moreover a convivial and entertaining companion. It was to the latter qualities that he owed his position at court under al-Muktafī and his successors, al-Muqtadir³⁸ (D. 320/932) and ar-Rāḍī (D. 329/940). To this last caliph we owe a happy reference to aṣ-Ṣūlī’s play. In his youth the chess-player had acted as his tutor, and a warm friendship seems to have arisen as a result. Al-Maṣ‘ūdī, who himself was intimate with aṣ-Ṣūlī and owed to him much of his information about the later caliphs, says (ed. cit., viii. 311: also h. Khallikān; and H, f. 13 b, where it is attributed in error to al-Muktafī):

It is related that ar-Rāḍī-billāh was once walking in his country seat at Thurayya, and called attention to a lovely garden, replete with lawns and flowers. He asked his courtiers if they had ever seen anything more beautiful. The courtiers immediately began to dilate on the wonders of the garden, to extol its beauty, and to place it above all the wonders of the world. ‘Stop,’ cried the caliph, ‘Aṣ-Ṣūlī’s skill at chess charms me more than these flowers, and more than all that you have mentioned.’

After ar-Rādī’s death, aṣ-Ṣūlī found himself out of favour, and an incautious statement that revealed his leanings towards the party of the ‘Alids (later the Shi‘ites) was so resented that he had to flee from Baghdād and go into hiding at Baṣra. Here he died in very reduced circumstances in 335/946.³⁹

Aṣ-Ṣūlī’s reputation in chess remained unchallenged in Arabic circles for more than 600 years. To his successors he represented all that was possible in chess, much as Philidor stood for the unattainable ideal to the early nineteenth century. His biographer, b. Khallikān says:

He stood alone in chess in his own time, for there was no one in that age who was his equal in skill. His play has passed into a proverb, and when men speak of any one who is remarkable for the excellence of his play, they say, ‘He plays chess like aṣ-Ṣūlī.’

Many Muslim players supposed from this proverb that aṣ-Ṣūlī was the actual discoverer or inventor of chess, and aṣ-Ṣafadī, b. Khallikān, and b. Sukaikir all point out the erroneousness of this belief.

We possess in the MSS. which have come down to us sufficient of aṣ-Ṣūlī’s work to form an opinion of the chess-activity of this master. We see him criticizing his predecessors not unkindly but with the touch of superior knowledge. We have his favourite openings, founded no longer on mere caprice but on definite principles. We have End-games which happened to him in play over the board and in blindfold play, with an occasional anecdote that shows how much the master’s play excelled that of his opponent. We see him as the first player to try to discover the science of the game or to enunciate the underlying principles of play. We may even possess some snatches of actual games in the analysis in the chess treatise contained in MSS. L and AE, the work of his grateful and able pupil al-Lajlāj.

This player, whose name is given by an-Nadīm in the Fihrist as Abū’l-Faraj Muḥammad b. ‘Obaidallāh, and in the MSS. as Abū’1-Faraj al-Muẓaffar b. Sa‘īd, probably owed his surname of al-Lajlāj (the stammerer) to a physical defect. The only fact that we know of his life is that recorded by an-Nadīm, who had seen him in Baghdād. In 360/970 he settled in Shīrāz at the court of the Būyid ‘Aḍudaddaula, where he died not long after. Both master and pupil are commemorated in a punning line in an elaborate essay in praise and dispraise of chess by Muḥammad b. Sharaf al-Qīrwānī, which is quoted by Hyde (ii. 57) from aṣ-Ṣafadī’s Sharḥ Lāmīyat al-‘Ajam.

Like aṣ-Ṣūlī, al-Lajlāj has been remembered as a great chess master, but while aṣ-Ṣūlī’s reputation has been in the main preserved in Syria and Egypt, al-Lajlāj’s memory has only survived among the Persians, the Turks, and the Moghul Hindus. To these peoples he has become the great historic figure in chess, and all the myths of the game have been attached to his name. As Lajāj, or more commonly Līlāj, he is the inventor of chess: he appears in the story of the Indian embassy to Nūshīrwān as the Indian ambassador; the fabulous Ṣaṣṣa b. Dāhir is represented as his father; and the Persian and Turk have forgotten aṣ-Ṣūlī entirely.⁴⁰

After the time of these great players there is a gap in the succession of references to chess at the court of Baghdād.⁴¹ The light of the Eastern caliphate was flickering out, and the centre of Muslim life was moving elsewhere. A few references may be quoted from other parts of Islam that show the wide spread of chess.

‘Omāra b. ‘Alī Najmaddīn al-Yamanī (D. 589/1175) in his Ta’rīkh al-Yaman (Yamau … by Najm ad-din ‘Omarah a1 Hakami, ed. H. C. Kay, London, 1892. pp. 88–92) gives a long account⁴² of the events leading up to Jayyash’s successful revolt at Zabīd in Southern Arabia in 482/1089. Jayyash had returned to Zabīd from India, and was living there in the disguise of an Indian faqir. He made use of his skill at chess to ingratiate himself with the vizier ‘Alī b. al-Kumm. To do this he took up his position each day at the bench at the outer gate of the vizier’s house.

Husayn, son of ‘Aly the Kummite, the poet, came forth on a certain day. He was at that time the most skilful chess-player of all the inhabitants of Zabīd. ‘Indian,’ he asked me, ‘art thou a good chess-player?’ I answered that I was. We played, and I beat him at the game, whereupon he barely restrained himself from violence against me. He went to his father and told him that he had been beaten at chess. His father replied that there had never been a person at Zabīd who could overcome him, excepting only Jayyash, the son of Najah, and he, he continued, has died in India. ‘Aly, the father of Husayn, then came forth to me. He was an exceedingly skilful player and we played together. I was unwilling to defeat him, and the match ended in a drawn game.

From this time Jayyash played frequently with the vizier, until he incautiously betrayed his identity by an involuntary exclamation after a game in which he had allowed Ḥusain to beat him for reasons of policy.

In Egypt the mad Fāṭimid ruler al-Ḥākim biamrillāh prohibited chess in the year 1005, and ordered all the sets of chess to be burnt. The order did not extend to the magnificent sets of chess in the palace treasury, for in a description of the treasures of a later ruler, al-Mustanṣir billāh (1036–94), al-Maqrīzī (D. 1441) mentions ‘chess and draught (read nard) boards of silk, embroidered in gold, with pawns (read men) of gold, silver, ivory, and ebony’. Much of this treasure had belonged to the ‘Abbāsid caliphs before the Fāṭimids acquired it.⁴³

I have already quoted from the Persian writer al-Bērūnī. His patron, the Ziyārid Qābūs b. Washmgīr (976–1012) of Tabaristān,⁴⁴ refers to chess in a poem in which he recounts his favourite occupations:

The things of this world from end to end are the goal of desire and greed,

And I set before this heart of mine the things which I most do need,

But a score of things I have chosen out of the world’s unnumbered throng,

That in quest of these I my soul may please and speed my life along.

Verse and song, and minstrelsy, and wine full flavoured, and sweet,

Backgammon, and chess, and the hunting-ground, and the falcon and cheetah fleet;

Field, and ball, and audience-hall, and battle, and banquet rare,

Horse, and arms, and a generous hand, and praise of my lord and prayer.

B. al-Athīr (Cairo ed., ix. 128) tells a story of the famous Maḥmūd of Ghaznī, which shows him as a chess-player.⁴⁵ In the spring of 420/1029 he seized Rai and dethroned Majdaddaula. He summoned the latter before him, and the following colloquy took place:

‘Hast thou not read the Shāhnāma and aṭ-Ṭbarī’s history (i.e. Persian and Arabic history)?’

‘Yes.’

‘Your conduct is not as of one who has read them. Do you play chess?’

‘Yes.’

‘Did you ever see a Shāh approach a Shāh?’

‘No.’

‘Then what induced you to surrender yourself to one who is stronger than yourself?’

Thereupon Maḥmūd exiled him to Khurāsān.⁴⁶

References to chess in Muhammadan Spain have, perhaps, a greater interest for us. I have already mentioned a chess poem by the Spanish poet ar-Ramādī as being quoted in the MS. H. There are one or two references belonging to the eleventh century. B. Hayyān (D. 469/1075), one of the best historians of Spain, records that the vizier Abū Ja‘far Aḥmad b. al-‘Abbās of Almeria, (D. 1038) was a keen chess-player. B. Ammār is said by al-Marrākoshī (writing 621/1224) to have played chess with the Christian King Alfonso VI of Castile, c. 1078. The poet b. al-Labbān ad-Dānī (c. 485/1092) wrote:

In the hand of fate we resemble the chess, and the shāh is often defeated by the baidaq.

There is a reference to chess in al-Maqqarī’s (D. 1041/1632) Nafḥ aṭ-ṭīb (ed. Dozy, &c., Leyden, 1855–61, i. 480) in connexion with the biography of the qāḍī Abū-Bakr b. al-‘Arabī (D. 543/1148); and b. Abī Uṣaibi‘a (D. 668/1270) in his K. ‘uyūn a1 anbā’, in his biography of b. Zuhr al-Ḥafīd of Seville (D. 596/1200), of Jewish descent, describes him as a good chess-player, who used to spend many an hour at chess with a friend of the tribe of al-Yanaqī.⁴⁷

Towards the end of MS. H (f. 51 a) we have a note from Abū’l-‘Abbās b. Juraij (B. 533/1139, D. 630/1232) that gives some information as to the chief players of his day in Spain. It will be seen that the blindfold game had many exponents.

Abū’l-‘Abbās b. Juraij said: I was contemporary with aṣ-Ṣaqālī (the Sicilian), al-Yahūdī (the Jew), and b. an-Nu‘mān, all of whom played blindfold: he goes on to say that Abū-Bakr b. Zuhair was equal to b. an-Nu‘mān. He says that Abū-Bakr b. Zuhair told him as follows: ‘There were assembled at one time in my house in Seville the following experts, aṣ-Ṣaqālī, his father, as-Sijilmāsī, aṭ-Ṭarābulusī, b. an-Nu‘mān, and az-Za‘farān.’ Abū’l-‘Abbās said: now in our time Muḥammad al-Ghamārī (?), Abū’l-Ḥusain b. ash-Shāṭibī, b. ‘Ulāhim al-Mukānisī, and Abū Muḥammad ‘Ahdalkarīm, an eminent man of Fez, formed one class, and b. Abī Ja‘far al-Mursī (the Murcian), b. al-Qaiṭūn, and b. Ayyūb and b. Abī’ẓ-Ẓafar b. Mardanīsh (?) formed another.

We have an interesting collection of players here from Sicily, Fez, Sijilmāsa, Tripoli, Murcia, and Seville. It evidences the spread of the game of chess in Muhammadan lands.

Both Fouché of Chartres⁴⁸ and William of Malmesbury,⁴⁹ in their accounts of the siege of Antioch (1097–8) during the First Crusade, tell how Peter the Hermit found the Turkish general Karbuga at chess when he was sent to treat with him at a critical point of the siege.

Aṣ-Ṣafadī (D. 764/1363), in his Sharḥ Lāmīyat al-‘Ajam, to which I have already referred, gives some interesting particulars as to chess in his day.

I once saw a soldier named ‘Alā’addīn in Egypt who was blind, and yet he used to play chess with the nobles and to beat them utterly. I say moreover that nothing pleased me more than the way in which he sat with us and talked and recited poetry, and narrated strange histories, showing that he was taking part in what we were doing. He would withdraw, and when he returned he had forgotten nothing that he had been doing. This is certainly surprising. The man was very famous in Cairo, and there were very few chess-players who did not know him.

At another time, in 731/1331, I saw in Damascus a man named an-Niẓām al-‘Ajamī, who played chess blindfold before Shamsaddīn. The first time that I saw him playing chess, he was playing with the shaikh Amīnaddīn Sulaimān, chief of the physicians, and he defeated him blindfold. We indeed knew nothing until he gave him checkmate with a Fīl, and we did not see that it was mate until he turned to us and said, ‘It is checkmate.’ I have also been told that he sometimes played two games at once blindfold. The sahib al-Maula Badraddīn Ḥasan b. ‘Alī al-Ghazzī told me that he had seen him play two games blindfold and one over the board at the same time, winning all three. He also vouches for this: Shamsaddīn once called to him in the middle of a game, ‘Enumerate your pieces (qiṭ‘a), and your opponent’s,’ and he rehearsed them in order at once, just as if he saw them before him.⁵⁰

Chess again appears under royal patronage at the court of the great Moghul emperor Tīmūr (B. 1336, D. 1405). His historian b. ‘Arabskāh (D. 854/1450) makes several references to chess in his ‘Ajā’ih al-maqdūr fī nawā’ib Tīmūr.⁵¹

Tīmūr ordered a city to be built on the farther bank of the Jaxartes, with a bridge of boats across the river, and he called it Shāhrukhīya.⁵² It was built in a spacious position. The reason why he gave this name of Shāh Rukh⁵³ to his son, and also to this city was as follows. He had already given orders for the building of this city on the river’s bank, and he was engaged in playing chess with one of his courtiers as was his wont: one of his concubines was also with child. He had just given shāh-rukh (check-rook) by which his adversary was crippled and weakened, and while his adversary was in this helpless position, two messengers arrived. One announced the birth of a son, and the other the completion of the city, and therefore he called both by this name⁵⁴ (i. 218).

Tīmūr was devoted to the game of chess because he whetted his intellect by it, but his mind was too exalted to play at the small chess (ash-shaṭranj aṣ-ṣayhīr), and therefore he only played at the great chess (ash-shaṭranj, al-kabīr), of which the board is 10 squares by 11, and there are 2 jamals, 2 zurafas, 2 ṭali‘as, 2 dabbābas, a wazīr, &c. A diagram of it is attached. The small chess is a mere nothing in comparison with the great chess ⁵⁵ (ii. 798).

Among chess-players (in Tīmūr’s reign) were Muḥammad b. ‘Aqīl al-Khaimī and Zain al-Yazdī, &c., but the most skilled at that game was ‘Alā’addīn at-Tabrīzī, the lawyer and traditionist, who used to give Zain al-Yazdī the odds of a Baidaq and beat him, and b. ‘Aqīl the odds of a Faras and beat him. Tīmūr himself, who subdued all the regions of the East and the West and had given mate to every sultan and king, both on the battle-field and in the game, used to say to him, ‘You have no rival in the kingdom of chess, just as I have none in government; there is no one to be found who can perform such wonders as I and you, my lord ‘Alī, each in his own sphere.’ He has composed a treatise on the game of chess and its situations. There was no one who could divine his intention in the game before he moved. He was a Shāfi‘ite…. He told me that he had once seen in a dream ‘Alī, the Commander of the Faithful, and had received from him a set of chess in a bag, and no mortal had beaten him since then.⁵⁶ It was noteworthy about his play that he never spent time in thought but the instant his opponent made his move after long and tedious thought, ‘Alī played without delay or reflection. He often played blindfold against two opponents, and showed by his play what his strength would have been over the board. With the Amir (Tīmūr) he used to play at the great chess. I have seen at his house the round chess (shaṭranj muddawara) and the oblong chess (shaṭranj ṭawīla). The great chess has in it the additional pieces that I have already mentioned. Its rules are best learnt by practice; a description would not have much value (ii. 872).

Wo have sundry references to this great master under the name of Khwāja ‘Alī Shaṭranjī in Persian literature,⁵⁷ while the MS. RAS gives no less than 21 positions from his games. When this circumstance is considered in connexion with the preface to this work, it certainly lends colour to the view that the MS. is the work which b. ‘Arabshāh tells us ‘Alī himself wrote. The passage runs—

I have passed my life since the age of 15 years among all the masters of chess living in my time; and since that period till now, when I have arrived at middle age, I have travelled through ‘Irāq-‘Arabī, and ‘Irāq-‘ajamī, and Khurāsān, and the regions of Māwarā’n-nahr (Transoxiana), and I have there met with many a master in this art, and I have played with all of them, and through the favour of him who is Adorable and Most High I have come off victorious.

Likewise, in playing blindfold, I have overcome most opponents, nor had they the power to cope with me. I have frequently played with one opponent over the board, and at the same time 1 have carried on four different games with as many adversaries without seeing the board, whilst I conversed freely with my friends all along, and through the Divine favour I conquered them all. Also in the great chess I have invented sundry positions, as well as several openings, which no one else ever imagined or contrived.

There are a great number of ingenious positions that have occurred to me in the course of my experience, in the common game, as practised at the present day; and many positions given as won by the older masters I have either proved to be drawn, or I have corrected them so that they now stand for what they were intended to be. I have also improved and rendered more complete all the rare and cunning stratagems hitherto recorded or invented by the first masters of chess. In short, I have here laid before the reader all that I have myself discovered from experience as well as whatever I found to be rare and excellent in the labours of my predecessors.

Chess remained one of the favourite recreations at the courts of Tīmūr’s descendants, and the Baber Autobiography (tr. Leyden and Erskine, London and Edinburgh, 1826, pp. 187–195), names several courtiers at the court of Ḥusain Mirza, King of Khurāsān (D. 1506), as inveterate chess-players. Among these were Zūlnūn Arghūn, Ḥassan ‘Alī Jelāir, Mīr Murtāz, and the poet Bināi of Heri.

My last authority for the unreformed Muhammadan chess is b. Sukaikir, the author of one of the MSS. which I have described in Chapter X. By birth a Damascene, he travelled through Syria, and visited Constantinople before filling the post of Preacher of the Mosque al-‘Ādilīya at Ḥalab (Aleppo), where he died 987/1579. In his chess work he mentions some experiences of his own. In 964 or 5/c. 1557 he saw a blind player at Damascus, who had played in the presence of the Sultan Sulaimān in Stambul. During the game the Sultan removed one of his men. The blind man quickly detected the fact, remarking that if the Sultan had done it there was nothing to be done but to play his best, but if any one else had done it he would appeal to the Sultan. In 967–8/1559 one of the best players in Damascus was a certain az-Zain al-Mathaka‘a. Once when he was on the point of mating an Egyptian, to whom he had given the odds of the firzān, a ragged Persian who was watching the game interposed and showed the Egyptian the move to thwart the attack. Az-Zain was naturally angry, and his anger was not appeased by the Persian telling him not to lose his temper because he did not know how to play. However, he agreed to play the beggar, who began by deliberately sacrificing faras, fīl, and three baidaqs without any equivalent. Then he asked az-Zain to name the piece with which he would choose to be mated. Az-Zain chose a baidaq, and the Persian gave him mate with a baidaq. Az-Zain, recognizing the Persian’s skill, took him into his service. The Persian would never play except at the odds of the ‘marked piece’. In 970/1562 he saw a Greek, Yūsuf Chelebī, at Trablis (Tripoli in Syria). This man used men of a larger pattern, and played blindfolded by touch. Finally he saw a blindfold player in Constantinople in 975/1567, who played often in his presence with uniform success. Like an-Niẓām al-‘Ajamī, he could at any time describe the position of every man on the board. The MS. Y narrates that there have been players who could play four or five games simultaneously blindfold, and goes on to say:

I have seen it written in a book, that a certain person played in this manner at ten boards at once, and gained all the games, and even corrected his adversaries when a mistake was made (Bland, 24).

Table of Contents

A variety of stories.—The oldest versions associated with India.—The connexion with nard.—The earlier legends from the chess MSS., al-Ya‘qūbī, al-Maṣ‘ūdī and Firdawsī.—The dramatis personae.—The story of the reward for the invention.—The Geometrical progression in literature.—Later stories introducing Adam, the sons of Noah, &c., and Aristotle.

The main facts of the earlier history of chess were well recognized by the older Muslim historians and chess-writers. They admit without reservation that the ordinary chess on the board of 64 squares was originally an Indian game which had reached them through the medium of Persia. But they were not content to leave the history in so bare a dress, and they endeavoured to take it farther back, to find a motive for the invention of the game, and to explain the manner of its discovery. Only in all this they had no historical foundations upon which to build other than the obvious relationship in arrangement, plan, and nomenclature that existed between the game of chess and the army and the tactics of war. This left an excellent opportunity for the literary artist, and he did not hesitate to adorn the story with details derived from his own imagination. Thus there appeared in quite early Muslim times a number of stories, more or less plausible, to account for the invention of chess, and the compilers of works on chess, from al-‘Adlī down, were diligent in collecting these from the sources at their service. Even writers of repute like al-Ya‘qūbī (c. 297/907) and al-Maṣ‘ūdī found a place for them in the pages of their historical works, while Firdawsī gave literary shape to one of the most widely known in the Shāhnāma. We find single legends repeatedly also in MSS. of miscellaneous contents in Arabic, Persian, and Turkish.

When we survey the material¹ at our disposal we find that the legends fall into three groups: those which are connected with India, those which associate the game with characters drawn from Scripture history, and those which bring in noted names from Greek philosophy. These two last groups are of later date, and have none of the detail that accompanies the stories of the first group, and it is not difficult to see a motive for the departure from the earlier association with India.

The legends of the earlier group are all, openly or tacitly, concerned with an Indian king or with the wise men of India. The connexion, however, is quite general, in that a special kingdom or district of India is seldom specified. The earlier Muslim writers appear to have formed their conception of a country on the model of the Eastern Roman Empire, or of the Sāsānian Empire which their forefathers had overturned.² India was to them a single kingdom, and it was long before they discovered that India was a geographical, not a political entity. Only a few of the legends give names to the king or sage of whom they treat, and still fewer attempt to fix the date at which the events they are recording took place. The ordinary story is quite indeterminate as to locality, dramatis personae, and date.

Several legends, however, connect the invention of chess in some way or other with the game of nard (tables, backgammon). We have already met with one instance of this association of games in the story of the introduction of chess into Persia in the time of Nūshīrwān. This linking of two games that to us seem so dissimilar—chess, a game in which chance plays the smallest of parts, and nard, a game in which chance plays the dominant part—appears somewhat singular, yet no association of games has been so persistent or has endured so long. It was not only prominent in Muslim lands, where it runs all through the legal discussion, the literature, and the traditions, but even in Christian Europe chess and tables appear in constant juxtaposition. The player of chess appears almost everywhere in the literature of the Middle Ages as a player of tables also, and the larger European problem MSS. treat of chess, tables, and merels. In these collections, however, the essential distinction between chess and tables is minimized, since in most of the problems on tables the constraint of the dice has been replaced by the liberty to select the throw desired, but this is, so far as the evidence goes, a purely European innovation. In Muslim literature it is upon the essential difference between chess as the game of skill and nard as the game of chance that stress is everywhere laid. The player’s complete liberty to select the move he wished to make in chess is contrasted with the player’s subjugation to the dominion of blind chance in nard. Throughout the legends with which I am about to deal, nard appears as the older and chess as the younger game; this is the reverse of what we find in the Nūshīrwān story as told in the Chatrang-nāmak and in the Shāhnāma. There, it will be remembered, the invention of nard is Buzūrjmihr’s reply to the Indian challenge to discover the nature of chess.

One of the older legends which occurs in AH (f. 1 b), C (f. 1 b), and V (f. 2 a), with the omission of all proper names, as an extract from the work of al-‘Adlī, and in almost identical words, with the addition of the proper names, in the Ta’rīkh of al-Ya‘qūbī (ed. Houtsma, Lugd. Bat., 1883, i. 99–102), brings the two games together. In this legend, an Indian monarch named Hashrān is represented as appealing to an Indian sage, Qaflān by name, to devise a game that should symbolize man’s dependence upon destiny and fate, and depict the way in which these forces work by means of man’s environment. The philosopher accordingly invented the game of nard, and explained to the king that the board stood for the year. It had 24 points (‘houses’) in all, because there are 24 hours to the day. It was arranged in two halves, each with 12 points to symbolize the 12 months of the year, or the 12 signs of the Zodiac. The number of men (‘dogs’) was 30, because there are 30 days to the month. The two dice³ stood for day and night. The faces were arranged with the 6 opposite to the 1, the 5 opposite to the 2, and the 4 opposite to the 3, so that the total of the dots on each pair of opposite faces should be 7, to bring in the number of days of the week and the 7 luminaries of the heavens.⁴ The players threw one of the dice in order to determine the order of play, and the one who secured the higher throw commenced, and moved his men in obedience to the throws given by the two dice. In this way man’s dependence upon fate for good or evil fortune was made evident. Hashrān was delighted with the game and introduced it in India, where it became extremely popular.

At a later date there arose a king, Balhait by name, who was advised by a Brahman that this game was contrary to the precepts of his religion. The king accordingly planned to replace nard by a new game, that should demonstrate the value of such qualities as prudence, diligence, thrift, and knowledge, and in this way oppose the fatalist teaching of nard. His Brahman friend undertook the task, and invented chess, explaining its name of shaṭranj by the Persian hashat-ranj, in which hashat means eight and ranj means side.⁵ The board was 8 by 8 squares, and there were 16 men (kalba, = dogs) on either side, viz. shāh, firz, 2 fīls, 2 faras, 2 rukhs, and 8 pawns. It was made on the model of war, because war is the most effective school for teaching the value of administration, decision, prudence, caution, arrangement, strategy, circumspection, vigour, courage, force, endurance, and bravery. Balhait was charmed with the game, and did his best to induce his subjects to adopt it in the place of nard.

Al-Maṣ‘ūdī’s version of the story is very similar, but there is some variation in the characters of the story. He does not, however, give it as one story, but places the two incidents in what he considered to be their proper chronology. Thus in ch. vii of his Murūj adh-dhahab (ed. cit., i. 157), under the reign of al-Bāhbūd, the eldest son of al-Barahman, we read:

It was at this time that nard and its rules were invented. It is symbolical of property, which is not the reward of intelligence or strength in this world, just as possessions are not gained by scheming. Others say that Ardashīr b. Bābak discovered and invented this game, which was suggested to him by the contemplation of the changes and caprices of fortune. He made its points 12 after the number of the months, and the men (‘dogs’) 30, after the number of days in the month. The two dice represent fate and its capricious dealings with men. The player, when the chances are favourable, secures what he wants; but the ready and prudent man cannot succeed in gaining what a happy chance has given to the other. Thus it is that property is due in this world to a fortunate chance.

A little later in the same chapter (ed, cit., i. 159) we read:

The next king (to Dabshalim) was Balhait. At this time chess was invented, which the king preferred to nard, because in this game skill always succeeds against ignorance. He made mathematical calculations on chess, and wrote a book on it called Taraq jankā,⁶ which has continued popular among the Indians. He often played chess with the wise men of his court, and it was he who represented the pieces by the figures of men and animals, and assigned them grades and ranks. He likened the Shāh to the chief ruler, and similarly with the rest of the pieces. He also made of this game a kind of allegory of the heavenly bodies (the 7 planets and the 12 zodiacal signs), and dedicated each piece to a star. The game of chess became a school of government and defence; it was consulted in time of war, when military tactics were about to be employed, to study the more or less rapid movements of troops. The Indians ascribe a mysterious interpretation to the doubling of the squares of the chessboard; they establish a connexion between the First Cause which soars above the spheres and on which everything depends, and the sum of the square of its squares. This number equals 18,446,744,073,709,551,615 … The Indians explain by these calculations the march of time and of the ages, the higher influences which govern the world, and the bonds which link them to the human soul. The Greeks (al-Yūnānīyan), the Byzantines (ar-Rūm), &c. have special theories and methods about this game, as we may see in the works of the chess-players from the most ancient down to aṣ-Sūlī and al-‘Adlī, the two most famous players of our time. Balhait reigned until his death, for 84 or, as other authorities say, 300 years. [His successor was Qūrush.]

The same legend, but told more baldly and with omission of names, occurs in Man., f. 16 b. The root idea of the story is seen in the witty remark which al-Maṣ‘ūdī quotes on a later stage of the same book (ed. cit., viii. 320), at the close of some additional remarks on nard.

Lastly, a Muslim philosopher has maintained that the inventor of chess was a mu‘taẓilite believer in the freedom of the will, while the inventor of nard was a fatalist who wished to show by this game that man can do nothing against fate, and that the true wisdom is to mould one’s life in agreement with the decrees of chance.

It is assumed in this legend that nard was a game of Indian invention,, and in so far the story is opposed to the other tradition, that nard was the invention of Artakhshīr the son of Bābakān, the first of the Sāsānian kings of Persia (A.D. 226–40), which is quoted at length in BM f. 5 b, in H f. 4 b, and in Man. f. 16 a. The attempt was made by later writers to bring the two legends into harmony by introducing modifications into the chess story. The motive for the discovery of chess is no longer the moral improvement of the Indian nard-players, but becomes the humiliation of the Persians. King Balhait is represented as being so aggrieved at the boastings of the Persians because of their discovery of nard, that he called upon a philosopher of his court, Ṣaṣṣa b. Dāhir, to invent a game that should transcend nard. The game of chess was Ṣaṣṣa’s reply. We find this in the chess MSS. H (f. 5 a), and Man. (f. 16 a)—in the latter from b. Taimīya (D. 728/1328).

It is this story which is included in the life of aṣ-Ṣūlī the chess-player in the K. wafayāt al-a‘yān of b. Khallikān (D. 681/1282),⁷ whence it was taken by aṣ-Ṣafadī (D. 764/1363) in his Sharḥ Lāmīyat al-‘Ajam, and by b. Sukaikir (S f. 25 a).

I have met many people who thought that aṣ-Ṣūlī was the inventor of chess. This is a mistake, for chess was invented by Ṣiṣṣa b. Dāhir for King Shihrām. Ardashīr b. Bābak, the founder of the last Persian dynasty, discovered nard, which was hence named nardashīr. Balhait was King of India at that time, and Ṣiṣṣa invented chess for him. The wise men of that time held it to be more excellent than nard. It is said that when Ṣiṣṣa had invented chess and produced it to King Shihrām, the latter was filled with amazement and joy. He ordered that it should be preserved in the temples, and held it the best thing that he knew as a training in the art of war, a glory to religion and the world, and the foundation of all justice. He expressed his joyful thanks for the favour which heaven had granted to his reign through such a discovery, and said to Ṣiṣṣa, ‘Ask whatever you desire,’ &c.

There is an obvious contradiction in this allusion, and both of the later writers endeavoured to remove it. Aṣ-Ṣafadī omits all mention of Shihrām, and names the Indian monarch Balhait throughout. B. Sukaikir, on the contrary, calls the monarch Shihrām and expressly describes him as an Indian king. He adds the note: ‘Some say that it was invented for Balhith, e.g. al-Yāfi‘ī.’⁸

The analogy existing between chess and war is the motive for four legends which are peculiar to the chess books. In one of these (BM f. 4 a, H f. 6 a, and RAS) the game is invented to find a distraction for a king who was passionately fond of war, but who had overcome all his enemies and was falling ill from ennui at not being able to pursue his favourite occupation. A philosopher produced for him chess, and showed him how he could still conduct forces and devise tactics in this game. The king tried the game, ascertained that the philosopher had spoken truly, and found distraction and health in playing chess. All the MSS. place the scene in India, H has no names for the characters of the story, BM calls the philosopher Ṣuṣa b. Dāhir, while RAS names the king Kaid, and the philosopher Ṣaṣṣa, placing the event shortly after the invasion of Alexander the Great. In this particular version, however, Ṣaṣṣa merely abridges the ‘Complete Chess’⁹ by reducing the size of the board from 11 by 11 to 8 by 8 squares, and the number of pieces from 56 to 32, because the Indians were incapable of appreciating so complicated a game. The complete chess itself was the invention of a Greek sage, Hermes, and had been introduced into India by Alexander and his soldiers.

In the second of these (AH f. 3 a, C f. 4 a, V f. 5 a, from al-‘Adlī’s work, and in RAS) the game is invented to assist in the military education of a young prince who pleaded that he was incompetent to lead his armies in war owing to his want of experience. The game of chess is alleged to have given the necessary training in tactics to convert him into an efficient commander. In both manuscript accounts the scene is laid in India, but RAS alone attempts to determine the characters of the story. These are stated to have been the young son and successor of Fūr (Pauras, the opponent of Alexander) and his vizier Ṣaṣṣa b. Dāhir. RAS again substitutes the abridgement of a Greek game for the invention of a new one.

The third story again represents chess as invented for the purpose of affording an opportunity for the practice of military tactics, and only differs from the previous legend in the matter of the particular circumstance of the invention. This story occurs in AH f. 3 b (C f. 4 b, Vf. 5 b, H f. 6 a, Man. f. 15 b, Y and S) as one of the three versions occurring in al-‘Adlī’s book. Its special interest consists in the fact that the game is represented as invented for a certain king¹⁰ named Shahrām¹¹ by the Indian sage Ṣaṣṣa b. Dāhir, who gave the game to the king ‘with the 14 ta‘bīyāt which are depicted in this book’.¹² The story of the reward is attached in AH to this story. B. Khallikān, at the end of his biography of aṣ-Ṣūlī the chess-player, interpolates a reference to this story when he mentions Shihrām as the monarch for whom Ṣiṣṣa b. Dāhar invented chess.

The fourth story is told in Man. f. 15 a, on the authority of b. Makhsharī. It is to the effect that a certain King of India, who was peaceably inclined, procured the invention of chess in order that his fellow-monarchs might settle their disputes over the board without effusion of blood.

I have left to the last what is probably the oldest of all the legends on the subject, dating back to pre-Muhammadan days. I have already called attention to the allusion to it in the tradition connecting the caliph ‘Omar b. al-Khaṭṭāb with chess, which I believe to be a genuine tradition. The legend is neither in al-‘Adlī (AH) nor in al-Maṣ‘ūdī, but al-Ya‘qūbī has a version of it which is interesting because of some of the details (ed. cit., i. 102–5).

It is related by some of the wise men of India that when Ḥūsīya, the daughter of Balhait, was Queen, a rebel rose against her. Now she was a prophetess with four children, and she invested her son. And the rebel slew her son. Now the men of her kingdom honoured him, and they guarded against her learning it. So they went to the philosopher Qaflān who was possessed of knowledge, wisdom, and prudence, and told him of it. He asked for three days and they granted it. He spent the time in thought. Then he called his disciple, ‘Summon a carpenter with wood of two colours, white and black.’ Then he devised the chessmen and ordered the carpenter to carve them. Next he called to him, ‘Bring me tanned leather.’ He ordered him to mark 64 squares on it, and he did so. Then he arranged a side, and studied it until he understood and had learnt it. Then he said to his disciple, ‘This is war without bloodshed.’ So he came to the men of the kingdom and produced it, and when they saw it they knew that no one exceeded him in wisdom. He made his disciple fight, and there befell shāh māt, and the Shāh was conquered. Now the Queen was interested in the news about Qaflān, and she visited him and bade him show her his invention. He called his disciple with the chess, and arranged it square by square. They played, and the winner said shāh māt. And she remembered and knew what he wished her to know, and she said to Qaflān, ‘My son is dead.’ He said, ‘You have said it.’ Then she said to the doorkeeper, ‘Let the people enter to comfort me.’ And when she had made an end, she summoned Qaflān and said to him, ‘Ask what you will.’ He said, ‘Give me a gift in grains of corn upon the squares of the chessboard. On the first square one grain (on the second two), on the third square double of that on the second, and continue in the same way until the last square.’ She said, ‘How much is this?’ and she ordered the corn to be brought. So they went on until she had exhausted the corn in the country. Then he estimated its value in money, and received that. And when this went on for a long time, he said, ‘I have no need of it: a small portion of worldly goods suffices me.’ Then she asked him about the number of grains that he had demanded.

Whereupon follows the total of the Geometrical Progression, which I give below.

There is a brief allusion to this story in H f. 6 a, but it is best known through its inclusion in the Shāhnāma (ii. 2889–3431; in Mohl’s edition, Paris, 1868, vi. 400–444),¹³ where Firdawsī names as his immediate authority a certain Shāhūī. As Nöldeke has pointed out,¹⁴ this is probably a misreading of the name Māhūī, Māhūī Chorsēdh, the son of Bahram of Shāpūr, being one of the four Zoroastrian priests to whom Abū Manṣūr al-Ma‘marī entrusted the work of arranging the national annals of Persia in A.D. 957–8. The section now bears the title of ‘The history of Gau and Talkhand, with the invention of chess’. The titles of the various sections do not, however, go back to Firdawsī.

The story treats of some incidents in the history of a kingdom in North-West India, which comprised Kashmīr and all the land to the confines of China, with Ṣandalī for capital. A king of this realm, Jamhūr, who excelled Fūr (Pauras) in fame, had died, leaving a widow and an infant son, Gau. He was succeeded by his brother, Mai, who married the widow and, after a short reign, died, leaving an infant son, Talkhand, who was five years younger than his half-brother. During the minority the widow held the regency, the question of the ultimate succession being left in abeyance. Each of the princes considered that his claim was the stronger, and their mother foolishly encouraged each in turn. As the boys grew up, the disputes became more bitter, and Talkhand adopted a most aggressive attitude. Gau, on the other hand, was as conciliatory as possible. Finally, however, Talkhand forced an appeal to the arbitrament of war. Gau gave the strictest instructions to his supporters that Talkhand’s life was to be spared. In the first battle Gau was successful, but Talkhand managed to collect his scattered forces, and a second battle took place close to the sea-shore. At the close of the battle, Talkhand was separated from his army and surrounded by the forces of his opponent, but when these came up to him, he was found to be already dead. The tidings plunged his mother into the deepest sorrow, and in her grief she accused Gau of slaying his brother. Gau defended himself, but to no purpose, and finally he offered to destroy himself if he could not demonstrate clearly to her how Talkhand’s death really happened. In order to compass this, Gau took counsel with his tutor, and by his advice convened all the wise men of the kingdom and laid the case before them. After a whole night’s consideration,

These experienced men ordered ebony to be brought, and two strong men made from it a square board to represent the ditch, the field of battle, and the armies drawn up opposite one another. They marked on this board 100 squares on which the armies and the two kings were to move, and finally they made two armies of teak and ivory, and two kings with heads erect, majestic and crowned. The infantry and cavalry formed the ranks in the battle array. They carved the figures of horses, elephants, viziers, and brave men charging on horseback against the enemy, all just as they went to the battle, some leaping in their haste, others moving calmly.¹⁵ Ready for battle, the Shāh (king) stood in the centre; on one side was the Firzāna (counsellor), his faithful companion. Next to the Shāh on both sides were two Pīls (elephants) who raised a dust, dark as indigo, about the throne. Two Shuturs (camels) were placed next to the Pīls, and two men of pure intention were mounted on them. Next to the Shuturs were two Asps (horses) with their riders, ready to fight on the day of battle. As warriors the two Rukhs at the two ends of the lines of battle raised their empty hands to the lips, as if to drink the foe’s heart’s blood. In front and rear moved the Piyāda (foot-soldiers), who were to come to the assistance of the others in the battle; if any pressed through to the other end of the field of battle, he was placed beside the Shāh like the Firzāna. The brave Firzāna never moved in the battle more than one square from his Shāh. The mighty Pīl ran through three squares, and observed the whole battle-field, two miles wide. The Shutur also ran through three squares, snorting and stamping on the field. The Asp’s move also extended over three squares, in crossing which one of the squares remained untouched. To all sides ran the vindictive Rukh, and he crossed the whole field of battle. Each piece moved in its own area, and made neither less nor more than its appointed move.¹⁶ If any one saw the Shāh in the battle, he cried aloud, ‘Remove, O Shāh!’ and the Shāh left his square until he was able to move no longer. The other Shāh, the Asp, Rukh, Farzīn, Pīl, and Piyādas had closed the road to him. When the Shāh had looked about him on all four sides, and with knit brows had seen his army overthrown, and his road barred by the water and the ditch, while the enemy were to left and right, before and behind, he died (was mate) of weariness and thirst.

Gau took this game of chess which thus explained the death of Talkhand to his mother. She continued to study it day and night without desiring food, until death released her from her sorrow. And from that time the chessboard has remained in the knowledge of mankind.

It is somewhat remarkable that in this legend Firdawsī has replaced the ordinary chess by a variety requiring an enlarged board, when no motive for the change can be discovered. As will be seen from the account of the derived forms of chess in Chapter XVI, he has not even adopted the standard variety on the 10 by 10 board of the chess books, but describes a form that is not mentioned elsewhere. The legend is repeated in RAS, as from the Shāhnāma, but the author of that MS. set out with the deliberate intention of enhancing the age and importance of another modification of chess, the Complete chess that was preferred by his sovereign the Mongol Tīmūr, and he has substituted for Firdawsī’s account of the invention a new version which makes Ṣaṣṣa b. Dāhir abridge the Complete chess into a game on the 8 by 8 board. We have already seen that he has dealt similarly with two other older legends.

Gildemeister (cf. Qst., 16) has expressed dissatisfaction with the ordinary texts of the Shāhnāma for this story. He points out that there is much variety of text in accessible MSS., and suggests that a scribal error first led to the appearance of the camel in one line which gives the names of the pieces, and that then later scribes restored the self-consistency of Firdawsī’s description by altering the dimensions of the board from 8 by 8 to 10 by 10, and introducing the lines relating to the camel’s position and move. It is much to be desired that a critical examination of the known MSS. could be made, but the immensity of the task of doing this for the Shāhnāma has probably deterred scholars from attempting it. The gain would not be worth the toil, except for points like the present, which do not touch the literary or historical value of the epic.

There is, however, at least one other work which makes the same substitution of the 10 by 10 board for the 8 by 8. This is the short history of ar-Ristāmī (840/1436–7), contained in MS. Gotha Arab. 1738 (old 1419). It mentions the introduction of chess into Persia thus (f. 3 a)—

After the sage Barzūya had brought the K. Kalīla wa Dimna from India with the Complete chess (ash-shaṭranj at-tamma), which has 10 by 10 squares, he translated it from Indian into Persian.

To this, however, I attach no importance. I do not know what authorities this late writer followed.

Various attempts have been made to identify the characters whose names recur most frequently in these legends, on the assumption that the names are really Indian in origin. The task is, however, one of great, if not insuperable, difficulty. The history of India, as it appears in the pages of early Muslim writers, is as unreal as their knowledge of the condition of India in their own days. Foreign names were peculiarly liable to misrepresentation when they were put into an Arabic dress. Moreover we are not certain of the forms of the proper names in the legends.¹⁷ The reader will have already noticed how I have used different vowels with different MSS. In the older Arabic MSS. the short vowels are unmarked, and when MSS. began to contain instructions as to the vocalization of the names, it was too late for them to have any historical authority behind, and the directions are based upon the analogy of native Arabic words. How unsafe a guide this analogy could be, we have already seen in the substitution of shiṭranj for shaṭranj. But there are other elements of uncertainty and error that are more serious still. The accuracy of the consonants in Arabic depends upon the close and accurate copying of the diacritical marks which distinguish many of the letters. Errors were always possible, but they are most dangerous in the case of foreign words, where detection is most difficult. If, again, the word has been derived from Pahlawī MSS., as is not impossible in the case of some of these legends, there is the additional possibility of error due to the deficiencies of the Pahlawī script. Nöldeke¹⁸ sees nothing impossible in tracing Shihrām or Shahrām, al-Ya‘qūbī’s Hashrān, the Dabshalīm of the Kalīla wa Dimna, and the Dewasarm of the Chatrang-nāmak all back to one Pahlawī original. If this be so, how can we feel certain of anything?

Among other suggestions as to the identity of Shahrām are Hyde’s (ii. 60), that the name is a scribal error for Baharam or Bahram, a name which occurs frequently among the Sāsānian kings, and also was used in India; and Pertsch’s, that Shahrām = Shāh Rāma (v. d. Linde, ii. 441). Sir H. M. Elliott in his History of India by its own historians (i. 409–10) suggests that Shahrām was Shahr Irān or Shahriyār (i.e. Kobād Shīrūyah), one of the last of the Sāsānian kings of Persia, who ruled for a few months (A.D. 628–9) during the disturbed period that followed the death of Khusraw II Parwīz. He, however, assumed that b. Khallikān described Shihrām as a Persian king, which is not the case. In any case it is difficult to see why the least important of all the Sāsānians should have been selected to adorn the legend. I return to Elliott’s argument below.

Balhait, Balhīt or Balhīth is the other Indian king who is frequently mentioned in the stories. Hyde (ii. 62) says that the form Balhīb also occurs. He suggested that these forms, which in the Arabic only differ in the diacritical dots to the last consonant, are intended to represent the Indian dynasty of the Balabhi or Balhara, who ruled in Guzerat from A.D. 319 to 613. This would make the name a title and not a personal name, and in this way he explains the apparent contradiction in the legend as given by b. Khallikān. This is ingenious, but not convincing, since other Arabic writers frequently use the correct form Balhara. It is, however, the only close resemblance that I can discover. Al-Maṣūdī’s succession of Indian kings—Barahman, 366 years; al-Bāhbūd, 100 years; Ramāh, 150 years; Fūr, 140 years (the Pauras or Porus of Alexander’s time, B.C. 326); Dabshalim, 120 years; Balhīth, 80 or 300 years; Kūrush, 120 years (who was followed by many Princes down to al-Ballahra, who was al-Maṣ‘ūdī’s contemporary in A.D. 943)—is of no assistance whatever to the solution of the difficulty.

Although no light has been thrown on the name Qaflān, the more ordinary name given to the inventor himself, viz. Ṣaṣṣa b. Dāhir,¹⁹ appears to be satisfactorily explained. These two names occur in connexion with a Brahman dynasty which ruled in the lower Scinde towards the close of the Umayyad caliphate, when the Muhammadans conquered this part of India. The kings of this family were Khakha, 632–72; Khandar, 672–79; Dāhir, 679–712. Khakha, the founder of the dynasty, appears in Persian histories as Chach the son of Silāīj, and in Arabic histories (aṭ-Ṭabarī, and al-Balādhūrī) as Ṣaṣṣa, while his son Dāhir retains his Indian name. Al-Balādhūrī gives the latter a son Ṣaṣṣa b. Dāhir, but only mentions him incidentally as having fled from the Muslims to a certain fortress. Elliott,²⁰ who develops the identification, is inclined to see more in it than a coincidence or a conscious appropriation of names. He thinks that the king Khakha or Ṣaṣṣa was the cause of the introduction of chess to the Western world, and associates in the story the nearly contemporary Sāsānian Shahriyār (Shīrūyah). I do not think that this view can be made to harmonize with the history of the game as now known. It puts the introduction into Persia too late for the facts, it ignores the difficulties that Shahrām in the stories is an Indian, not a Persian king, that Ṣaṣṣa is the son, not the father, of Dāhir, that Ṣaṣṣa is a philosopher, not a usurping monarch. I think the truth is to be found in the view that the earliest teller of the legend chose the Indian names that were most familiar to his generation, in order to give verisimilitude to his story. This leaves to Khakha the more modest share in the history of chess of lending his name to the hero of chess-romance.

Bland (62) suggested that Ṣaṣṣa is a corruption of the name Xerxes, and identified him with the philosopher who in European fable is associated with the discovery of chess.²¹ I am inclined to agree with his identification, only I think the perversion of name has been in the other direction, and that the European Xerxes is an attempt to explain the Arabic Ṣaṣṣa.

All the MSS., al-Ya‘qūbī, and b. Khallikān add to one or other of their legends a conclusion which tells how the philosopher was rewarded for his invention of chess. When the king invited him to choose his own reward, he is said to have asked for a quantity of corn which was to be placed upon the chessboard in a particular way. The first square was to hold one grain,²² the second two, the third four, the fourth eight, and so on, each square containing double the number of grains that were placed upon the preceding square. The quantity of corn asked is, of course, enormous, the number of grains being the sum of a geometrical progression of sixty-four terms, with 1 for the first term and 2 for the common ratio. The total is 2⁶⁴ –1, or

18,446,744,073,709,551,615 grains,

a quantity which would cover England to a uniform depth of 38·4 feet.²³ It is added that the king did not know which to admire the most, the invention of chess or the ingenuity of the request.

This calculation is undoubtedly of Indian origin, the early Indian mathematicians being notoriously given to long-winded problems of this character. In its earliest form it may be older than chess, and be based upon the ashṭāpada board.²⁴ I have already quoted a passage from al-Maṣ‘ūdī in which he speaks of the importance which the Indians attached to the sum of the Progression. It would appear to have also been a favourite calculation among the Muslims, though they generally shirked the complete solution by reducing to larger units whenever the figures grew inconveniently large. This also made the immensity of the sum more easy of comprehension. Thus al-Bērūnī reduces the total to 2,305 mountains ‘which is more than the world contains’. He also makes use of the real sum in his Al-āthār al-bāqiya (ed. Sachau, Leipzig, 1878, and Eng. tr., London, 1879) to illustrate the different systems of numeration current in his day. At least two Arabic treatises were written on the problem, viz. the Taḍ‘if buyūt ash-shaṭranj of al-Missisī (9–10th c. A.D.)²⁵ and the Taḍ‘if‘adad ruq‘a ash-shaṭranj of al-Akfānī (D. 749/1348), and several shorter discussions occur in MSS. which I have seen. The MS. Man. gives no less than five methods of treating the problem, one from b. Khallikān (who naïvely states that he did not believe the total could be so great until he met an accountant of Alexandria who showed him the actual calculation), two from ar-Rāghib,²⁶ the fourth from the Durrat al-muḍī’a of Quṭbaddīn Muḥammad b. ‘Abdalqādir, and the fifth from al-Akfānī. MS. Gotha Ar. 1343 has also three calculations, the last of which is interesting since the story is different from the usual one. In this a Sultan who used to challenge all comers at chess, beheading all whom he defeated, after beating ninety-nine opponents met his superior in a dervish. The latter claimed the usual reward—in dirhems.

The calculation reached Europe with the Arabic mathematics, and was discussed by Leonardo Pisano in his Liber Abbaci. Other European references to the Progression will be found below in Part II, Chapter IX.

The later Arabic legends which bring chess into association with Bible history need not detain us. They are clearly an attempt to rehabilitate the game of chess at a time when the legal schools were looking with disfavour upon it. The earliest record of this type of tradition that I know occurs in the preface to aṣ-Ṣūlī’s K. ash-shaṭranj. After referring to al-‘Adlī’s statement that chess was invented by Ṣaṣṣa b. Dāhir, aṣ-Ṣūlī goes on to say that this is a fabrication which he had found in many works. For himself he preferred to accept the ‘statement based on sound tradition’ which he traces back to Ka‘b al-Akhbār, one of the most notorious forgers of traditions that Islam ever knew, that chess was invented by Būshāqūs, Yūsh‘a b. Nūn (Joshua) and Kālab b. Yūfannā (Caleb), and that the first who played the game was Qārūn (Korah). Būshāqūs taught the game to the Persians. Later writers are still more daring in their assertions. The MS. H suggests that chess was invented by Adam to console himself for the death of Abel, and numbers Shem, Japhet, and King Solomon among the chess-players.

From the time of al-Ma’mūn onwards, the writings of the more famous Greek philosophers became known to the Muslim world in translation. It was, perhaps, inevitable that the scattered allusions to the Greek board-games which occur in Plato and other writers should be misapplied to chess, but to this we owe the statements in H and later chess books that Aristotle, Galen, and Hippocrates were also chess-players.

Table of Contents

The chessboard.—The names of the chessmen in Muslim lands.—Symbolism of the game.—Forms of the chessmen.—The arrangement of the men for play.—The moves of the chessmen, and technical terms.—Relative values of the pieces.—Aim and method of play.—Notation.—Concordant and discordant men.—Classification of players.—Gradations of odds.—Etiquette of play.

The shaṭranj board resembles all native Asiatic boards in being unchequered, but differs from the Indian and other boards in showing no trace of any regular marking of certain squares. The term ‘board’, however, is somewhat deceptive. The Arabic names,¹ ruq‘a (a patch or piece of paper), sufra (a table-cloth or napkin), naṭ‘a (a cloth) and bisāt (a carpet), all imply a soft material, and from the earliest days of the Muslim game down, the board has generally been a square piece of cloth or other substance upon which the dividing lines of the squares (Ar. bait, house, pl. buyūt; Per. khāna, pl. khānahā; Turk, au, pl. aular) are worked in another colour. In more elaborate chess-cloths the individual squares may bear a pattern of some simple type, or be merely indicated by the regular recurrence of a conventional design which occupies the centre of the otherwise undivided squares, while these patterns or designs may even, as in the ease of the so-called Turkish cloth of which Falkener gives a photograph (196), show a further differentiation on lines analogous to the Indian marked squares. In the desert rougher materials still are employed: Stamma (Noble Game of Chess, London, 1745, xii) notes:

The wild Arabs draw the Squares on the Ground, and pick up Stones of different Shapes and Sizes, which serve them for Pieces.

Boards of more solid materials—it will be remembered that al-Ya‘qūbī describes Qaflān as making his board of leather—and even chequered boards are not entirely unknown, but the chequering is incidental to the ornamentation of the chessboard, and is not essential for its use. With the fondness of the Egyptian, the Turk, and the Persian for inlaid work in wood, it would be strange indeed if so obvious a method of beautifying the board had not suggested itself. The artist, painting a chess scene for some MS., found the same device at hand.² But all these are by way of exception only; with the limited powers of move of the older Muslim game, the chequered board was less of a convenience than it is in modern days, when more pieces move with larger sweeps.

Other Arabic terms in connexion with the chessboard which I have noted from the older MSS. are ḥāshyā, margin (generally of the Rooks’ files); wijh, wujh, jiha (pl. jahāt), jānib, side or wing of the board; nāḥīa al-firzān, the Queen’s wing; zāwaya, rukn (R and H only), corner square; wasṭ ar-ruq‘a, the four central squares; sāf (pl. sāfāt), file, as sāf ar-rikkākh, the Rooks’ file; ṣaff (pl. ṣufūf), ṣaffa, file or row, as in ṣaff al-awsaṭ, a central file; mashya (rarely), file.

The names of the chessmen (Ar. dābba, pl. dawābb, beast—used apparently at times in a more restricted sense, e.g. L, 14 b, firzān wa dawābb, Queen and Pieces, and f. 65 b, dawābb kull wa bayādiq, all the Pieces and Pawns; qiṭ‘a, piece; kalb, pl. kilāb, dog; mithāl, pl.’amthila, tamāthīl, figure. Per. kālā, pl. kālāhā, in RAS only; muhrah. Collectively: Ar. ālāt ash-shaṭranj) that are used by the Muhammadans of India and Malaya have been already given, but it will be convenient to collect in a table those that are used by other Muslim peoples. To these I add the Abyssinian (Amharic) terms, since the Abyssinian chess is a variety of the Muslim game.

(NOTE. The ordinary Arabic names are those of the MSS.; the other terms, which I designate as colloquial, are taken from descriptions by Europeans: Hamilton, who gives houssān from Egypt, and Grimm, who obtained his terms from Syria.)

The military character of chess was well understood by the earlier Arabic writers on the game. Apart from many allusions in general literature, there are three descriptions in the chess MSS. which I quote because of the light which they throw upon other features of the Muslim game. The first of these is the work of aṣ-Ṣūlī, and is contained in AH (f. 19 b), V (f. 12 b), and Man. (f. 27 b):

The chessmen are classified in this chapter. The shāh, it is said, is the king. The firzān is the vizier, because he protects and covers the king, and is placed next to him, advancing before him in the battle. Muḥammad b. ‘Abdalmalik az-Zayyāt³ says, ‘How beautiful is the function of the fīl in chess! He resembles the secretary who reveals and plans. His use in war is slight except when he does a deed of renown. His is the secretary’s cunning, as when he gives shāh-rukh or shāh-faras or forks two pieces. Or, perhaps in another game when a number of pieces are collected against him, and he draws the game since none of them can attack him. The firzān has the same power. In a case like this the fīl is better than the faras, and when there are several firzāns it is even better than a rukh when the latter cannot attack it.’ The faras, it is said, is different: he is a bold horseman, and this is his function in chess. The rukh, it is said, is like a commander and a general of an army: like the faras he is a horseman, and the command is his. His work is to confine the game, and his strength is manifest when the ninth (read seventh) of the game is his.⁴ The baidaqs (Ar. bayādiq), it is said, are like the foot-soldiers who move in advance and hinder the horses (’afrās) and rukhs (rikhākh); but when the rukh gets behind them and attacks them from the rear, he destroys them just as horsemen in war destroy the foot-soldiers.⁵

The second passage occurs in Man. (f. 18 a), in the course of a version of the Ṣaṣṣa legend:

The ḥakīm arranged it thus and it was chess, and he made it in the likeness of a battle between two armies. He made the nafs to resemble the king, the firzān to resemble the vizier, and the two rukhs the commanders of the right and left wings. Next to these he placed the two faras, and then the two fīls. These are reckoned as the more important members of the army. The baidaqs resemble the infantry. He then made each piece’s move (ḥaraka) proportionate to its strength in the army, and fixed that the victory is gained when the shāh is slain, his army still being in existence—this is mate—or when his army is captured.

In the third passage from BM (f. 6 b) a different interpretation is attempted:

The inventor of chess made the board to represent a field of battle upon which two armies are drawn up for the contest, and the six figures, shāh, firzān, fīl, faras, rukh and baidaq represent the six classes upon which war turns and which are essential to it. Of these the first is the king who rules, the second the vizier who leads, the third the commander of the army who arranges, the fourth the cavalry, the fifth the fortresses (ḥuṣūn), and the sixth the infantry. He represented the king by the shāh, the vizier by the firzān, the commander of the army by the rukh, the cavalry by the faras, the fortresses by the fīls, and the infantry by the baidaqs. This is the classification of the chessmen (ālāt ash-shaṭranj).

The following was his intention in the arrangement. He put the shāh in the centre because the king ought to be in the heart of his army. He put the firzān next him because the vizier ought to be in the king’s vicinity. He put the fīl next the firzān because the strongest places in the battle array ought to be where king and vizier are. He put the faras between fīl and rukh because cavalry ought to be the defence of the fortresses. He put the rukh next the faras because the commander ought to be in command of the right and left wings. He put the baidaqs in a line in front of these because the infantry is placed in the van in battle. This was his intention in the arrangement of the chessmen.

His intention in the arrangement of the pieces in the game was to liken the game to a struggle and attack. He gave the baidaq a move of one square in a straight line, because it is not right for the foot-soldier to quit his position in battle, nor to advance except step by step. He made it take obliquely because the injury he inflicts on his enemy in the battle happens unexpectedly. He appointed that he should become a firzān when he reached the end of the board, because a man who advances and penetrates to the enemy’s camp, and preserves himself from capture or overthrow, deserves the viziership in war. He gave the faras a far-reaching move, because the horseman, being mounted, can transport himself to a distance, and can fall back to his camp when he is threatened. He made his move an oblique one in moving forwards and backwards and in capturing, because the horseman of necessity attacks his enemy, lance in hand, and takes him by swift and sudden movements. He gave the rukh the move in the four cardinal directions as far as the end of the lines confronting him, which is the most extended move of the pieces, because in war it is the commanders of the right and left wings who harass and burden to their utmost the enemy’s weak points which are opposite to them. He made the shāh’s move a single square in every direction, because the King is not one who should move swiftly. He is free to move at choice either forwards or backwards. The rule for the firzān is the same, except that his liberty of move is less than the shāh’s. When he takes, he takes according to his ward.⁶

The text omits the description of the move of the fīl, but I have given it entire as in the MS. because it shows very clearly the extent to which the original parallelism that was intended between chess and war was still recognized by Arabic writers as late as 1250 A.D., notwithstanding the fact that the older Indian explanation had been forgotten. The explanation of the Rook as a commander shows that the original meaning of the name of the chess-piece was not generally known, while the new interpretation of the fīls in the BM extract suggests that the use of the elephant in war was also passing from memory. The new interpretation is far-fetched, and yet after all only a foreshadowing of the European substitution of the Castle (at first borne on an elephant’s back) for another piece, the meaning of whose name has been entirely forgotten, the Rook.

I have already cited passages to show that the use of pieces (ṣūra) which were actual images of the men and animals from which the chessmen took their names was opposed on religious grounds. The legal objection to so elaborate a type of piece was undoubtedly assisted by the economic difficulty that few players would be able to afford such costly implements of play. The invention of a simpler and cheaper type of chessmen was a direct result of the great popularity of chess, and at an early date a definite conventional type of man came into use. The oldest examples are the chessmen from Bambra-ka-thūl in the British Museum, but there are other early Muslim chessmen, mostly from Egypt and probably none as old as 1000 A.D., in the Museum in the case of chessmen in the Mediaeval Room. They may be easily identified, from their resemblance either to the modern Muslim chessmen or to the earlier European conventional pieces. We may safely conclude that the original Muslim type comprised (a) three pieces of different sizes, but all more or less cylindrical in shape, of which the tallest represented the shāh, the intermediate one the firzān, and the smallest (an exact replica of the firzān) the baidaq; (b) two pieces with long and narrow necks, one with a slightly cleft head for the fīl, the other for the faras; and (c) another piece, rather more massive than the fīl or faras, with a well-marked top which in early times was flattened on two sides and contained a deep cleft in the centre, which represented the rukh.

The opposing sides were distinguished by the different colours of the two sets of chessmen. In the MSS. these are called red and black (probably because inks of these colours were most easily procurable), and only rarely white and black. The modern sets which I have seen are white and black, white and red, red and green, and red and black.

The arrangement of the pieces at the opening of the game is generally shown in the MSS. as here diagrammed. In only one MS., the late Pers. Y, is the red King placed on e1. At first sight this appears to be out of harmony with the European arrangement. The latter is, however, purely conventional, and depends upon the convention governing the placing of the chequered board and the very modern custom of giving the first move exclusively to White. In earlier times the Black pieces were preferred:—H (f. 51b) says, ‘the Black men are for the chief, and the White for the inferior’; and the chess-player generally visualized the board from the Black point of view. The important fact in these MS. diagrams is the unanimity with which they support the European opposition of King to King and Queen to Queen, and oppose the antiquity of the modern Asiatic crosswise arrangement.

The moves of the pieces were as follows:—

The Shāh or Nafs, King (K),⁷ moved one square at a time into any of the eight or fewer squares surrounding that on which he happened to be standing, the square selected being unoccupied by one of his own pieces or a protected piece of his opponent’s, and being out of the range of attack of any hostile piece at the moment of moving. He captured in the same way that he moved. He could not move into check, and whenever he was checked he was obliged to remedy it as in the modern game. If he was unable to do so it was shāh māt (Per. shāh māt), māt, rarely shāh wa māt, checkmate, and the game was ended. When a player gave check it was usual for him to warn his opponent of the fact by saying shāh, coupling with the name of the King the name of any other piece that was simultaneously attacked, e.g. shāh war-rukh, shāh wa rukh (Per. shāh rukh), a check forking King and Rook; shāh wal-fīl, a check forking King and Fīl (L, f. 26 a); shāh wa firz (H, f. 37 a), or shāh wa firzān (AH, f. 55 b), a check forking King and Firzān; shāh wa faras (AH, f. 56 a), a check forking King and Faras; and even shāh wa baidaq ash-shāh (L, f. 63 a), a check forking King and King’s Pawn. Another technicality in AH (f. 92 b) is shāh mubaṭṭanān, an intimate check, used of a cheek by a Rook on an adjacent square (e.g. Re7 checking Ke8). From this use of the name of the piece is derived the verb shāha (imp. yashīhu; IV stem, ’ashāha; VII stem, inshāha), to check. At a later date it became usual to say kisht (also written qish, qishāh) and kisht māt instead of shāh and shāh māt when giving check and checkmate, and this is the ordinary expression at the present time in Arabic, Persian, Turkish, and Urdū. The earliest examples I know occur in the Turkish MS. F. For discovered check we have the technicalities in Arabic shāh fīl-kashf (kashf discovery) and shāh min i‘irā (shāh min al-i‘ra), where i‘rā (from the root ‘ariya, to be naked) is a term peculiar to chess, occurring in Persian as ‘irā, in Urdū as ‘arop, and in Malay as aras. The Persian Madar al-afāzil (Bland, 49) defines ‘irā as ‘that piece at chess which is interposed between a King and a Rook to protect’, but in the Arabic MSS. it is used rather of the whole position of a file dominated by a Rook, in which the check is for the moment covered by an intervening piece of either colour between the Rook and King. We have accordingly such expressions as ‘to move into i‘rā’ (to play the King on to a file where there is the possibility of a check by discovery by the removal or capture of an intervening piece), ‘to expose to i‘rā’, ‘the position in i‘rā’. To cover check is in Arabic satara ash-shāh; to mate, māta (imp. yamūtu).

The Firzān (pl. farāzīn), firz, or firza, Counsellor (Q),⁸ also moved one square at a time, moving diagonally into any one of the four or fewer diagonally adjacent squares to that on which he was posted, the square chosen being unoccupied by a piece of his own side. He captured in the same way that he moved. He could only play to 32 squares, and on a chequered board he would be restricted to squares of one colour. Al-Lajlāj attached great importance to the development of the Firzān in the game, and aimed at securing a clear path (Ar. sabīl or ṭarīq, pl. ṭurūq and ṭuruqāt) by which it could be brought into the opponent’s half of the board.

The Fīl (pl. fiyala, ’afyāl), or Elephant (B), possessed a diagonal move, which consisted of a leap over a diagonally adjacent square, whether occupied or empty, into the square beyond on the same diagonal. This is commonly, though ambiguously, described as a diagonal leap into the third square; it is a move familiar to English draughts-players as the move of a man in making a single capture without the removal of the man thus captured. The Fīl captured as he moved. Only eight squares of the board were accessible to any Fīl, and no Fīl could ever encounter or attack any other Fīl. The two Fīls were distinguished as Fīl ash-shāh, KB, and Fīl al-firzān, QB, or as the right- and left-hand Fīl. The King’s Fīl was also called Fīl al-qā’īma, Fīl al-man‘a, the drawing Fīl, or Fīl an-nafs, the Naf’s Fīl.

The Faras (pl. ’afrās), or Horse (Kt), and the Rukhkh (pl. rikhākh, mod. rikhakha), or Rook (lit. Chariot) (R), possessed precisely the same moves as their European equivalents, Knight and Rook, possess to-day. The squares commanded by one of these pieces were termed in Arabic its muqāṭa‘a, or province. Other technicalities are jama‘a, to double, to place both Rooks on the same file; ar-rukhkh al-a‘la, the front Rook of two on the same file.

The Baidaq, Baidhaq (pronounced baizaq; pl. bayādiq or bayādhiq), or Pawn (P), moved and captured as the European Pawn does, with the difference that it possessed no power of moving over two squares for its first move. There is consequently no question of one Pawn taking another in passing. On reaching the eighth line it ceased to be a Pawn, and was at once promoted to the rank and took the name of Firzān. No other promotion was possible, and there was no limit to the number of Firzāns that a player might possess at any time of the game. The Arabic verb to promote, ‘queen’, is farzana (V stem tafarzana). The Pawns were distinguished by associating with them the name of the piece on whose file they were standing, e.g. baidaq ash-shāh, King’s Pawn, &c. In addition, the Rook’s Pawn was often called baidaq al-hāshīyā (hawāshī), the marginal Pawn, and the King’s or Firzān’s Pawn, baidaq aṣ-ṣadr, the central Pawn. Fanciful names were attached to the advanced Firzān’s Pawn (baidaq as-saif, the sword Pawn), and the advanced King’s Bishop’s Pawn (baidaq as-sayyāl, the torrent Pawn) in the analysis of the opening developments. Other terms that I have noted are baidaq al-firzān al-’aṣlīya, the original QP of two on that file; and baidaq al-faras aṣ-ṣadr, the front KtP of two on that file; and baidaq firzān al-aswad al-mutaqallab ‘an baidaq shāh-hu, Black’s QP that has been changed from KP (by making a capture). Al-Lajlāj attaches great importance in his analysis to the maintaining of an advanced Pawn, and speaks of establishing (Ar. vb. makuna, V stem, tamakkana) a Pawn, of the establishment (tamkīn) of a Pawn, and of an established Pawn (baidaq tamkīn), meaning the posting of a Pawn on an advanced square, and its support there so that it was practically untakable except at the cost of superior force.

The Arabic MSS. which I have used supply chess uses of many ordinary words in connexion with the movements and other activities of the chessmen. Some of these may be noted here. To move a piece for the first time is kharaja (IV ’akhraja), to develop. Of ordinary moves the ordinary expression is, ‘White comes with his Rook to (jā’a, bi,’ila) such a square’: but this is only rarely used of the Fīl or Baidaq. A player pushed (dafa‘a) a Baidaq (very rarely also a Rukh, Faras, Firzān, or Shāh), and threw (ramā, also of Faras and Firzān; ṭaraḥa, also of Faras; or alqaya) or shifted (ḥawwala) a Fīl. He placed (waḍa‘a or ja‘ala), played (la‘iba), went away with (dhahaba bi, also of Faras), or advanced (madda) a Rukh. The Shāh and Firzān ascended (ṣa‘ada or ṭala‘a), descended (nazala or ḥadara, VII’inḥadara), or entered (dakhala, IV ’adkhala) a square. Or the Shāh removed (ba‘uda), passed (marra), walked (masha), or limped (zamala)—all suggested by his move of a single square. In general use I have noted ḥaruka (II ḥarraka), zāla (II zauwala, IV ’azāla), naḥa (II naḥḥa, V tanaḥḥa), and naqala, all meaning move, remove. For the substantive, move, there is similar diversity of expression. In addition to the general terms ḥaraka, ḍarb (pl. ḍarba, ḍurūb), mashya (walk) is used of the Shāh, taḥrīk or ḥurk of the Firzān, ṭarḥ and nazwān (leap) of the Fīl, munzā of Firzān, Fīl and Faras, daf‘or daf‘a of the Baidaq. Adverbs of direction are mustawīyan, in a straight line, farasīyan, as a Faras, firzānīyan, as a Firzān, and fīlīyan, as a Fīl. More general terms are ’akhadha, to take (’akhadha bāṭilan or majjānan, to take without loss); dhahaba, ‘atā’a or hāta bāṭilan, to sacrifice; ḍaraba or waqa‘a ‘ala, to attack; or wuqū‘, an attack; ḥāmala, to offer to exchange (of Rukh only); or ṣarafa, to exchange; waqa‘a baina, to fork; ḥabasa, ḥaṣara, to shut in or confine; ḥafiẓa, ḥamā, to defend; ḥifẓ., hamā, himā’i, defence.⁹

Both aI-‘Adlī and aṣ-Ṣūlī made an attempt in their chess-books to estimate the relative values of the chessmen in the early part of the game. The method adopted was based upon a monetary scale and the dirhem was taken as a convenient unit. BM (f. 11a) gives a brief extract from al-‘Adlī’s work, and AH (f. 14 b), V (f. 14 a), II (f. 18 a—shortened text), BM (f. 10 a), and Man. (f. 27 a) give aṣ-Ṣūlī’s chapter. An independent, but not materially different, estimation is given in AE and RAS. I have adopted the text in Man. as the basis of the following translation.

Values of the chessmen, calculated for their original positions.

Aṣ-Ṣūlī has said: The Shāh is reckoned beyond value because of his superior dignity. The highest in value after the Shāh is the Rukh. Its value is one dirhem. The Faras’ value is dirhem. The Firzān’s value is dirhem, but some say dirhem. The Fīl’s value is dirhem. KP and QP, each dirhem; BP and KtP, each dirhem rising to dirhem; a marginal P dirhem because it can only take on one side. I consider the KKtP better than QBP; aṣ-Ṣūlī has said that this is because this Baidaq is a spy against the return of the opponent’s stronger Fīl and Firzān. He gave these values for the commencement of the game; the values of the pieces may increase or diminish afterwards. The better of the Fīls is the Fīl an-nafs, which is the Fīl al-man‘a (the drawing Fīl, called by al-Lajlāj, AH f. 133 b, Fīl al-qā’īma). It is better than KP, and the other Fīl is better than QP. The two central Baidaqs are better than Firzān and Fīl, and any two Baidaqs are better than the Firzān. Rabrab and Abū’n-Na‘ām used to advise the exchange of the Firzān for the two marginal Baidaqs, the exchange of the Fīl for two Baidaqs under all circumstances, and the exchange of the weaker for the better Baidaqs. Do not exchange Fīl and Baidaq for Firzān unless your opponent has gained your Firzān. If your Rukh is confined, try to exchange it for Faras and Firzān, but otherwise not.

AH goes on to show that these values may be altered completely in the End-game, where even the advantage (Ar. faḍl) of a Rukh may be insufficient to convert a draw into a win, e.g. Kt v. R is a drawn game, and so is R and Kt v. R. Even the Fīl might become of higher value than a Faras (Kt v four concordant Qs loses, but B v. four concordant Qs draws) or a Rukh (R v. four concordant Qs loses, B v. the same draws).

The values in RAS are calculated with greater nicety, but the MS. shows a tendency to over-estimate the value of the minor pieces. The MS. gives Q or dirhem; B or dirhem; KP dirhem; QP dirhem; Kt and BP dirhem; and RP dirhem.

Aṣ-Ṣūlī’s estimate enables us to form some kind of comparison with the modern game. I adopt as the unit of my scale the value of the RP in the Muslim game, and as the connecting link the value of the Knight, whose move is the same in both games.

The great increase in the powers of the Bishop and Queen in modern chess has naturally resulted in a relative diminution in the value of the other pieces.

The method of play in the older Muslim game was identical with that followed in the modern European game. The players played alternately, each making a single move (dast, pl. dusūt; ḍarb) in his turn of play.¹⁰

The aim of play was twofold, either to give checkmate to the opponent’s King, or to annihilate his army. To this latter form of victory I have given the Middle-English name of Bare King, which answers closely to the terms Shāh munfarid, isolated King, and mufrad, pl. mufridāt, isolations, of the Arabic MSS. It was obtained whenever a player captured the whole of his opponent’s army, the King excepted, and still retained some of his own men upon the board, or at least one man out of reach of the opponent’s King. If the solitary King could take his opponent’s last remaining man in his move following that in which he was bared, the game was considered drawn. We have already seen (p. 57) that in Ḥijāz there was a local variation in the rule regarding Bare King. There a solitary King was defeated the instant that he was bared, whether he could bare his opponent the following move or not. This win was called the Medinese victory.¹¹

It occasionally happened in the course of the game that a player, whose turn it was to play, found himself unable to move any of his pieces in a legal manner, and yet at the instant his King was not in check. This Ending, to which we give the name of stalemate (Ar. zā’īd, rarely mazīd, from the verb zāda, imp. yazīdu; or zā’īda to stale, deprive of the power of moving, very rarely used of any other piece than the Shāh), was decisive in Muslim chess. The player who found himself in this predicament was held to have lost the game.¹²

A game to which for any reason a decisive result could not be obtained was said to come to a stand (Ar. qā’īm, ’iqāma, qayām; pl. qawā’īm, qā’īma, ’iqā-māt—all derived from qāma, to stand; cf. mod. qūwima, to be drawn), or to be inaccessible (mana‘a: a later term, not used in AH, frequent in H), i.e. to be drawn. This might happen from equality of force (e.g. R v. R); insufficiency of force (e.g. R v. Kt); inability to force exchanges (e.g. Qs and Bs moving on squares which on a chequered board would all be of one colour v. Qs and Bs moving on squares of the other colour); perpetual check; or repetition of move. The problems will contain examples of all these forms of drawn game.¹³

The chess MSS. employ two different methods of describing the squares of the board, which we may conveniently distinguish as the descriptive and the literal, or as it is often called, the algebraic notations. The former is by far the commoner, and is indeed employed in all the MSS. except Oxf. in the problem solutions.

In the descriptive notation the board is regarded as belonging half to one player and half to the other, and the two halves are called Red’s and Black’s accordingly. In each of these halves the squares are defined in terms (1) of the side, e.g. left-hand or right-hand, less frequently King’s and Firzān’s; (2) of the master-piece of the file; (3) of its distance from the edge where the player sits. Thus, assuming that Red sits at the foot, Black at the upper edge of the diagram, g4 is Red’s right-hand Faras’ fourth square, and g5 is Black’s left-hand Faras’ fourth square. Very occasionally the notation was extended right across the board, and we have g7 termed Red’s right-hand Faras’ seventh square. A little ambiguity is introduced as a result of the want of fixity in the initial positions of King and Firzān. In the analysis in L, where the original position of the Kings is known absolutely, the notation is consistent throughout, but in the problem solutions where it is impossible to say for certain where the King stood originally, now the e-file and now the d-file is called the King’s. In many problem positions, the Firzān’s file can be determined on the assumption that the Firzāns in the diagram are the original Firzāns of the game, and in these cases the solutions almost invariably name the central files accordingly. The important fact is that the notation in the vast majority of the solutions assumes the normal arrangement of the opposing Kings. In a few solutions (not more than five in all) the notation is confused, and squares on both central files are described as on the King’s file. I have only found one solution in which the notation is consistent with a crosswise arrangement of the Kings.

This notation does not possess the merit of brevity, but its clearness has given it a long and fruitful life, and with but slight modification it is still the most popular notation in Europe outside of Germany and Switzerland, and in America at the present time. Al-‘Adlī strongly advised players to commit it to memory, and it forms the foundation upon which Y built its hints upon playing chess blindfold.

In the algebraic notation each square is denoted by two letters, the first of which is common to all squares of the same file, the second to all squares of the same row. It is very similar to the notation which I employ in this book. Thus the successive files which I call a, b, &c. are termed in AH, where this notation is employed in connexion with the Knight’s Tours, t, sh, r, f, m, l, k, y, with the numerical meanings of 400, 300, 200, 100, 40, 30, 20, 10. In Oxf., where this notation is used in all the problem solutions, the letters f (80), ‘(70), s (60), n (50) are substituted for the older letters of the files a—d. Both MSS. use for the rows 1, 2, to 8, the letters a, b, j, d, h, w, z, h, with the numerical meanings 1, 2, 3, 4, 5, 6, 7, 8. Thus c5 is r h in AH or s h in Oxf., e3 is m j, f6 is l w, &c. This notation was also introduced into Europe—first at an early date, and again, in an improved form, by Philip Stamma of Aleppo in 1737. With small alterations it has become the normal notation of German chess-players.

F (Q) and R alone, of all the MSS. which I have consulted, make no use of a regular notation, but describe the moves by reference to other pieces on the board, e.g. ‘moves next to the Firz,’ or by such adverbs as ‘up’, ‘down ‘aslant’, ‘straight’.

The unchequered state of the Muslim board made it a matter of some difficulty to see whether two Firzāns on the same side could defend, or whether two Firzāns of different sides could attack one another. This was a matter of very considerable importance in the Ending, and might mean all the difference between a won and a drawn game. The older masters (see BM, f. 10 b) gave cumbrous roles by which a player could ascertain whether a Pawn promoted on a certain square would move on the same 32 squares as another Firzān. Firzāns and Fīls which moved on different sets of squares (different coloured squares on a chequered board) were said to be discordant (Ar. khālif, mukhtalif, takhāluf, mukhālif, contrary, different), while those which moved on the same set of squares (squares of the same colour on a chequered board) were called concordant (Ar. talāqī, mutlaqī, mutalliq, that which meets, or less frequently muwāfiq, concordant). The existence of these terms is clear evidence for the uncoloured nature of the Muslim chessboard.

Most of the MSS. attempt a classification of chess-players in regard to their skill in play. The different classifications do not entirely agree, and the discussion would seem from the first to have been more academic than practical. Probably at no time did a position in any but the highest class carry any great weight in popular estimation. Apparently al-‘Adlī was the first to treat of the classes of chess-players, but we only know of his proposals from a brief reference which aṣ-Ṣūlī added at the end of his own chapter on the question.

Al-‘Adlī recognized five classes of players. The highest contained the ‘ālīyāt or grandees. The second class, the mutaqāribāt or proximes, received the odds of the Firzān from the ‘ālīya. The third class received the Rook—‘which is silly,’ comments aṣ-Ṣūlī. We know nothing of the remaining classes.

Aṣ-Ṣūlī also recognized five classes, and gives the name of ‘ālīya to the highest. There have never been more than three at any time or place in this class. He names Jābir, Rabrab, Abū’n-Na’ām al-‘Adlī and ar-Rāzī as having been of the first class. The later MS., BM, substitutes al-‘Arī for ar-Rāzī (a clerical error, I believe) and adds the names of two Baghdād players, b. Dandān, and al-‘Qunāf, who must accordingly have flourished between 950 and 1250. The later MSS. merely repeat aṣ-Ṣūlī’s information, and strangely enough none adds aṣ-Ṣūlī’s own name, or that of al-Lajlāj, both of whom were certainly of the highest skill. Aṣ-Ṣūlī goes on to say that Rabrab and ar-Rāzī were the greatest of these masters, that al-‘Adlī had for a while stood alone in the class until ar-Rāzī challenged him and proved his mastership, that ar-Rāzī also stood alone in the class for a time and died before another grandee appeared. The second class, the mutaqāribāt or proximes, contains players who win from two to four games in ten when playing with a grandee, and who receive odds from him, the best, QKtP or RP. the weakest, KP or QP. The grandee is credited with the ability to calculate ten (AH says twenty—an error, surely) moves ahead, the proxime sees far less. The test of a player’s class is his success with a player of known class when playing without odds. If he wins regularly seven or more games in ten, he belongs to a higher class; if fewer, not. The third class receive the odds of the Firzān from the ‘ālīya, the fourth class the Faras, the fifth class the Rook. If a player requires greater odds his skill is accounted nothing; as a player he is beneath contempt. Al-‘Adlī once said, ‘You do not give the odds of the Faras to a player who can plan shāh wa rukh or shāh-māt.’ Elsewhere in the preface of his book, aṣ-Ṣūlī specially instances skill in solving manṣūbāt (problems), knowledge of the Endings and End-game decisions, and knowledge of the ta‘bīyāt, and when to abandon or modify them in play, as distinguishing marks of the ‘ālīya.

The later MSS., H and Man., recognize six classes, introducing one between the third and fourth of aṣ-Ṣūlī’s list, who receive odds from the ‘ālīya greater than the Firzān but less than the Faras.

Closely connected with the classification of players is that of the proper gradation of odds (Ar. ḥaṭṭ). The only discussion of this occurs in RAS, and Forbes (99) abridges the passage thus:

Having now explained the moves of the pieces and their exchangeable value, I shall proceed, O Reader! to inform you of the different degrees of odds established by the masters of old. A true Chess-player ought to play with all sorts of people, and in order to do so, he must make himself acquainted with his adversary’s strength, in order to determine what odds he may give or accept. A man who is unacquainted with the rules for giving or receiving odds is not worthy of the name of Chess-player. It is only by equalizing the strength of the combatants that both of them may reap amusement and edification; for what interest could a first-rate player, such as ‘Adali (i.e. al-‘Adlī), or Ṣūlī, or ‘Alī Shaṭranjī, find in playing even with a man to whom they could each give the Knight or the Rook?

The smallest degree of odds, then, is to allow the adversary the first move. The second degree is to give him the Half-Pawn, which consists in taking either Knight’s Pawn off his own file and placing it on the Rook’s third square. The third species of odds is the giving the Rook’s Pawn; the fourth, that of the Knight; the fifth, that of the Bishop; the sixth, that of the Queen. The seventh degree of odds is to give the adversary the King’s Pawn, which is the best on the board. The eighth species of odds is the King’s Bishop. The ninth is the Queen’s Bishop. The tenth degree of odds is the Queen. The eleventh, the Queen and a Pawn; or what is equivalent, a Knight; for though the Queen and Pawn be slightly inferior to the Knight at the beginning, yet you must take into account the probability of the Pawn becoming a second Queen. The twelfth species of odds is the Knight and Pawn. The thirteenth, the Rook. To give any odds beyond the Rook can apply only to women, children, and tyros. For instance, a man to whom even a first-class player can afford to give the odds of a Rook and a Knight has no claim to be ranked among Chess-players. In fact, the two Rooks in Chess are like the two hands in the human body, and the two Knights are, as it were, the feet. Now, that man has very little to boast of on the score of manhood and valour who tells you that he has given a sound thrashing to another man who had only one hand and one foot.

There is an interesting passage in H, ff. 50 b–51 a, in the middle of an anthology of poems relating to chess, which shows that it was thought useful to discuss the proper line of play to adopt when giving odds. The passage is not very clear, but it deals with the opening play when the odds of the Rook are given in return for a Pawn, the odds of Faras for Pawn, of Faras, of Firzān, and of a Pawn. In the first case, when the Rook is given for a Pawn, two lines of play are given, but it is not stated which Rook and which Pawn are to be removed from the board. As is often the case in Arabic analysis, the play on one side only is given; it can be taken as suggesting the lines upon which the player should attempt to model his development. The two lines of play are as follows: I. (1) Pe3; (2) Kte2; (3) Pg3; (4) Pg4; (5) Ktf4; (6) Ktg2; (7) Pf3; (8) Pf4; (9) Qe2; (10) Qf3; (11) Ph3; (12) Pd3; (13) Pd4; if the opponent now moves P(c6)c5, do not take the Pawn, but play (14) Pc3; if he takes the d-Pawn, play (15) eP×P; (16) Bd3; (17) Pf5. If he takes this Pawn, then retake by g-Pawn. II. (1) Pc3; (2) Pc4; (3) Pd3; (4) Ktc3; (5) Pb3; (6) Pd4; (7) Qc2; (8) Pe3; (9) Kd2; (10) Pa3; (11) Pa4; (12) Ba3; (13) Bd3.

The same line of play is recommended when giving the Faras for a Pawn, but is not advisable in the case of the odds of the Faras alone. The play when giving the Firzān is discussed in a single sentence, too corrupt to be intelligible. When giving a Pawn, the following plan of development is given as best: (1) Pd3; (2) Pd4; (3) Pc3; (4) Pf3; (5) Kth3; (6) Ktf2; (7) Qc2; (8) Qd3; (9) Pg3; (10) Ph3; (11) Ph4.¹⁴ If, however, the opponent play first,

move the Pawns in a body, and do not let him outstrip them. Then bring your Kt to e2. If he moves against your d-P, do not take him until he takes. If you take first, it is to his advantage and spoils your game. If he takes, it is not advisable to take with c-P.

The discussion is interesting, as showing that chess analysis was carried on in Muslim circles to a greater extent than had generally been supposed was the case.

The later MSS., and especially Y,¹⁵ attach considerable importance to the etiquette of play. Thus when two players sit down to their game, the lower in rank is to spread out the board, and to shake the pieces from the bag in which they are kept. He is next to wait until his superior has made his choice of colour, and in arranging his men he is to take care not to place his King and Firzān until his opponent has placed his; he is then to place his King opposite to the other King. If the players are of equal rank, the first to seize the men chooses the positions of the Kings. The stronger player should offer fair odds. Ordinary rules of good manners should be observed; onlookers should keep silence and refrain from remarks on the state of the game or from advice to the players. An inferior should not wilfully play to lose.

It is quite evident from the stories of the early Muslim players that much of this etiquette did not obtain in their time. Traditions respecting Sa‘īd b. al-Musayyib, al-Ḥaṣan al-Baṣrī, ash-Sha‘bī, and Muḥammad b. Sīrīn relate how these tābi‘s used to advise players as to their moves while watching the game.

Текст предоставлен ООО «ЛитРес».

Прочитайте эту книгу целиком, купив полную легальную версию на ЛитРес.

Безопасно оплатить книгу можно банковской картой Visa, MasterCard, Maestro, со счета мобильного телефона, с платежного терминала, в салоне МТС или Связной, через PayPal, WebMoney, Яндекс.Деньги, QIWI Кошелек, бонусными картами или другим удобным Вам способом.