^{page 24 note 1} Compare, for example, the various meanings of πεσσός or κυβός as given by Pollux.

^{page 24 note 2} See Murray, H. J. R., History of Chess, p. 161.Google Scholar

^{page 24 note 3} ‘This bore no stronger resemblance to Chess than a coal barge does to the Great Eastern’ (Forbes, , History of Chess, p. 184).Google Scholar

^{page 25 note 1} Hence the common identification with the Greek πόλεις παίʒειν is impossible. The fact that no dice were used should alone have prevented Isidore's information from being referred to latrunculi (see later).

^{page 26 note 1} L.L. x. 22 ‘ad hunc quadruplicem fontem ordines deriguntur bini, uni transversi, alteri derecti, ut in tabula solet in qua latrunculis ludunt.’

^{page 26 note 2} Thus it was desirable for a piece to be covered where possible by an adjacent piece; otherwise it could only avoid capture by a hard fight (Ovid, A.A. iii. 359; see Owen, S. G.'s note on Trist. ii. 479–80Google Scholar, where Ovid describes the same tactics in greater detail).

^{page 26 note 3} Cf. the lion in Fig. 1 (from a papyrus in the British Museum).

^{page 26 note 4} Falkener, , Games Ancient and Oriental, pp. 37 ff.Google Scholar, claims to reconstruct the rules, but his theories are quite fanciful and untenable, resting as they do on an unwarranted identification of latrunculi with the Egyptian game of Tau; further, he wrongly adduces Isidore, Orig. xviii. 67.

^{page 26 note 5} Cf. Murray, , op. cit., p. 30, n. 15Google Scholar. For the finds on the Wall see Archaeologia Aeliana, 1913, p. 62Google Scholar; Transactions of the Cumberland and Westmorland Antiquarian Society, 1913, p. 338Google Scholar, and 1926, p. 444 (references kindly given me by Miss A. S. Robertson of Glasgow University). For the Chedworth specimen see Transactions of the Bristol and Gloucestershire Archaeological Society, xlv, p. 285Google Scholar. For the Corbridge find see Fig. 2 (illustration from A Roman Frontier Post, by James Curle, with the kind permission of the author and of the publishers, Messrs. Maclehose).

^{page 27 note 1} The same method of capture obtained in a Persian game mentioned in Firdausi's Shahnama, and is found in similar games played to-day in Japan and Siam. Mr. Murray notes: ‘The old Norse game of Hnefatafl preserves the ‘latrunculi’ capture, but the game, as also the obsolete Welsh Tawlbwrd, became a hunt rather than a fight, a kind of Fox and Geese. But the older Welsh game of the Arthurian legend, and the old Irish game Fidchell (the same as the Welsh Gwydbwyll) seem to have been identical with the Roman game.’

^{page 28 note 1} A similar account is given by von der Lasa in Schachzeitung, 1863 (a series of articles on Greek and Roman games), and by Tilley, A. in C.R. vi, p. 235 f.Google Scholar

^{page 30 note 1} Tabula aperta does not imply a folding board, as some have suggested: the phrase merely corresponds to the military aperto campo.

^{page 30 note 2} Hampered as the writer is by the demands of metre and rhetoric, he speaks throughout as if a single man could effect a capture; actually, of course, the method of capture described by Ovid and Martial must have obtained; that the game for which Piso was so famous was latrunculi is expressly stated by the Scholiast to Juv. v. 109, but it is odd that the author of L.P. nowhere uses the word latro.

^{page 30 note 3} Experiment clearly illustrated this operation; ancipites means something like ‘enfilading’. Mora is probably used quite generally, and is not a technicality meaning ‘check’, as supposed by Becq de Fouquières and others.

^{page 31 note 1} Minor references are Cic. de or. i. 217, Quintil. xi. 2. 38, Val. Max. viii. 8. 2 (all relating to P. Mucius Scaevola, a famous player).

^{page 31 note 2} See illustration in Lafaye's article s.v. tessera, in Daremberg and Saglio, Dictionnaire des Antiquités.

^{page 31 note 3} See Owen, on Trist. ii 483Google Scholar illustration in Daremberg and Saglio, s.v. lusoria tabula, and in Boeswillwald, Cagnat, and Ballu, , Timgad, p. 20.Google Scholar

^{page 32 note 1} There is a large collection of these boards in Max Ihm's paper in Bonner Studien (Berlin, 1890), pp. 223 ff.Google Scholar, supplemented in Römische Mitteilungen, vi. (1891), pp. 208 ffGoogle Scholar. Their subjects fall into three main groups: (a) maxims for players, (b) jeers at ‘rabbits’ (cf. the first example quoted above), (c) references to the circus, eating, and other pleasures. Ihm does not relate them to xii scripta but to an admittedly similar game. See also C.I.L. viii. 7998Google Scholar, 8407, ix. 4907, x. 546, xiv. 5317; Dessau, , Inscr. Lat. Sel. ii. 8626Google Scholar; Journal of Roman Studies, xiii, p. 164Google Scholar; L'Année Épigraphique, 1930, No. 68.Google Scholar

^{page 33 note 1} I doubt if the passage in A.A. iii. 355–6, quoted by Owen as parallel and referred by him to the same (unknown) game, is really more than a prefatory couplet describing the application of dice to board games in general.

^{page 33 note 2} A 3 × 12 board for this Egyptian game is in the Boulak Museum at Cairo; see Falkener, , op. cit., p. 97.Google Scholar

^{page 33 note 3} Cf. also my forthcoming paper in J.H.S.

^{page 33 note 4} This board has a diagram on the other side which probably served for a simple type of merels (for which see later).