up and a fourth one picked up in the same way. The four are then tossed up and caught, the fifth stone being picked up while they are in the air. Fifth round. One stone is placed on the back of the outstretched hand. It is then tossed up and another is picked up before it is caught in the palm. While this is held in the palm, the other is tossed up from the back of the hand and a third one picked up, the falling stone being caught in the same palm. This is repeated with the other stones, those previously picked up being retained in the hand. A variant is played by tossing up a marble each time instead of one of the stones, and catching it after its rebound from the stone pavement. This use of the marble resembles the northern practice, but of course cannot have been the original mode of playing the game. I did not learn the Arab or Indian game. POL-KURU KELIYA, 'The Coconut-pins Game.' This is the game called Spelicans in England, into which country it was imported from Holland, according to Skeat's Etymological Dictionary. The Dutch may have learnt it in India, where it is well known. The thin splinters or 'spells' of ivory or bone with which it is sometimes played are carved in the form of various Eastern weapons-swords, bows, arrows, spears, as well as saws, trumpets, and some ancient military standards. These are dropped on a table in a promis !c cuous heap, crossing each other as much as possible, and are removed one by one by means of a small ivory hook, the aim of each player being to take out as many as possible without the slightest movement of the others, which disqualifies the player for that turn. The Sinhalese game is played in exactly the same manner. As the name implies, short lengths of the rib of the side leaf of the coconut, 6 or 7 inches long, thinned down and well smoothed, are used instead of splinters of bone or ivory. Their number varies. One set made for me numbers about 120, but I was informed that a full set should number 240 or 300, although so many are not often employed. On every tenth stick notches are cut on one edge according to the number of tens, one at the tenth stick, two at the twentieth, and so on. For drawing out the sticks a small hook is cut at the end of one of them, or the end is bent sharply round for the purpose. The turn to draw out the sticks passes round the players consecutively to the right, each one stopping when some movement is observed among the other sticks. When a stick is once touched by the hook, even by accident, the player must draw out that one, or attempt to do it. The game ends when all the sticks have been drawn out; and the winner is the player who can count the highest score. In this, one is counted for each notch on the sticks drawn out by the player, as well as for each unnotched stick. The aim of the players is therefore to acquire the sticks with the highest numbers of notches. NERENCHI KELIYA, written Niranchy' by Ludovisi.1 The meaning appears to be connected with the Tamil verb nirei, 'to fill up,' or 'become full,' and ansi (pronounced anchi), 'play.' FIG. 245. Nerenchi This is undoubtedly a very ancient game, the age of which is unknown. Perhaps the earliest evidence of it in Ceylon occurs at Mihintale, where two diagrams for playing it, called Nerenchi-peta, were cut on the great flight of steps, thirty feet wide, for ascending the lower part of the hill, by the masons who laid them. I have not met with any record of the construction of these steps. Tennent states that the monks at Mihintale informed him that the work is attributed to Mahādāṭhika Mahā-Nāga (9-21 A.D.).2 Forbes says the same. In the much worn inscription left by a King Nāga on a vertical rock near the Aet dāgaba, on the crest of the hill, 1 The Sports and Games of the Sinhalese,' by L. Ludovisi. (Journal of the Ceylon Branch of the Royal Asiatic Society for 1873, p. 17.) 2 Ceylon, 2nd ed., Vol. ii, p. 606, foot-note. P P already mentioned in the account of that structure, the reference to the gift of the steps at the dāgaba may be taken to prove that the much more needed flight at the lower part of the hill was already built by 245 A.D. In that case the most probable time for their construction was during the second century or at the beginning of the first century A.D., when extensive works were undertaken there. Two other diagrams cut on a rock near the Lankārāma dāgaba are illustrated by Mr. Bell; near them are some letters in 'rock character,' but whether pre-Christian or later Mr. Bell does not state.1 One diagram which measures 8 inches by 7 inches, is of the usual form; the other consists of two crosses, one vertical and one oblique, inside an oblong measuring 51 inches by 5 inches. A similar design to the illustration, 15 inches square, is also cut on one of the great slabs which roofed part of the temple begun by Rameses I (1400-1366, B.C.) and completed by Seti I (1366-1333 B.C.) at Kūrna, in Upper Egypt, on the western side of the Nile valley. Many other designs, which are illustrated in connection with the next chapter (Fig. 273), are on the same roof, and three of them certainly were incised before the stones were finally laid, since in trimming the edges of the slabs on which they occur so as to make them fit against the adjoining ones, the masons cut away part of these diagrams. If, as appears most probable, the persons who cut these damaged designs also made the rest, it is evident that the knowledge of this game must have been possessed by the ancient Egyptians in the fourteenth century B.C. This is the more likely since among the other diagrams on the same slabs there is an upright cross enclosed in a square, which the small holes marked at the angles and intersections of lines appear to show was used for playing a still simpler form of the game, that was like the 'Noughts and Crosses' of English children. Both in Cairo and at Luxor I was informed that the Nerenchi game is not known at the present day in Egypt, which is equivalent 1 Arch. Survey of Ceylon, Third Progress Report, p. 5, foot-note. to saying that it is not an Arabic game. Thus the diagram was not cut by modern Arabs. The diagram for the Nerenchi game in its simplest form consists of a plain cross enclosed in a square, or an open cross of double lines enclosed in a square or circle. Next, we have two intermingled crosses in a square,1 one being upright and the other diagonal. In Ceylon all these are favourite designs as charms against planetary and demoniacal influences. Lastly, there is the full design, which consists of a small central square resting on a cross, with two enclosing squares, the central square being further protected against evil influences by having a 'guarded' cross inside it. In the Sinhalese diagram for the game given by Mr. Ludovisi it is interesting to find a small plain cross drawn inside the central square, as in Egypt, where however the cross is a 'guarded' one, having a cross-bar at the end of each arm. The three parallel-sided squares, one inside the other, are also found on articles taken from European Lake Dwellings, where a cross is placed in the centre. They occur at the first city at Troy, without the central cross, and are to be seen on Indian punch-marked coins. They are included in my Sinhalese manuscript book of magical formulas as a diagram which guards against evil caused by planets and demons. It is probable that in early times this game may have been thought to have some mystical or magical significance. The mystical number three which recurs so often in it was connected with the early deities of Egypt and the Euphrates valley. The number three was also reckoned the first of the odd, or lucky, numbers 2; therefore to win a game in which victory went to the player who obtained the greatest number of this lucky figure may have been thought an auspicious omen. All idea of such a meaning is now unknown to those who play the modern game. 1 This form of board was employed for the game in Ireland in the early part of the 18th century, according to Col. Wood-Martin (Pagan Ireland, p. 536); and a stone counter and several bone disks which are thought to have been used for such a game have been found in Irish Lake Dwellings (op. cit. p. 534). 2 Plutarch's Morals. On the E at Delphi, viii. The simplest form of the game, the familiar Noughts and Crosses,' is not, I think, found in the interior of Ceylon, but the complete game is well known there and is also played in India, as well as in Europe. It is not known in Western Africa, nor have I found any references to it byAfrican travellers. I have also been informed that it is unknown in Japan. In Ceylon the diagram for it is drawn on the ground. The game requires two players who alternately lay down a small counter-usually a stone or fragment of earthenwareat one of the angles, or the points where the arms of the cross meet the sides of the squares. While doing so, on each occasion when a player forms a row of three of his own pieces, which is termed 'Nerenchi,' he lays down an additional piece. When only two places remain unfilled the next player moves one of his pieces into one of the vacant points, and the play is continued by the two players, who move their pieces alternately, each one endeavouring to form a row of three of his own pieces, which the other tries to prevent. Whenever a row is so formed the player who has obtained the Nerenchi removes an opposition piece from the board and has an additional move. The play ends when one player has lost all his pieces. HAT DIVIYAN KELIYA, The Game of the Seven Leopards.' This game is mentioned by Ludovisi, who gives a copy of the diagram on which it is played. This is an isosceles triangle with a central upright from the middle of the base to the apex, and two other lines across it parallel to the base and ending at the sides of the triangle. It is played by two persons, one of whom has one piece called, according to Ludovisi, the Tiger,' while the other has seven pieces called 'Leopards,' which are captured and removed off the board when the Tiger jumps over them one at a time, into an empty place. The Leopards win the game if they can shut him up or 'imprison' him so that he cannot move. The pieces move along the lines of the figure to all junctions of lines, going one step at a time except when the Tiger is making a capture. The Tiger is first placed at the apex of the |