fellows when the hours arrive for relieving guard; the game is never stopped for more than a couple of minutes at a time, viz., when the cards run out and have to be re-shuffled. This brief interruption is commonly considered to portend a break in the particular vein which the game may have happened to assume during the deal--say a run upon black or red, an alternation of coups (in threes or fours) upon either colour, two reds and a black, or _vice versa_, all equally frequent eccentricities of the cards; and the heavier players often change their seats, or leave the table altogether for an hour or so at such a conjuncture. Curiously enough, excepting at the very commencement of the day's play, the _habitues_ of the Trente et Quarante tables appear to entertain a strong antipathy to the first deal or two after the cards have been "re-made." I have been told by one or two masters of the craft that they have a fancy to see how matters are likely to go before they strike in, as if it were possible to deduce the future of the game from its past! That it is possible appears to be an article of faith with the old stagers, and, indeed, every now and then odd coincidences occur which tend to confirm them in their creed. I witnessed an occurrence which was either attributable (as I believe) to sheer chance, or (as its hero earnestly assured me) to instinct. A fair and frail Magyar was punting on numbers with immense pluck and uniform ill fortune. Behind her stood a Viennese gentleman of my acquaintance, who enjoys a certain renown amongst his friends for the faculty of prophecy, which, however, he seldom exercises for his own benefit. Observing that she hesitated about staking her double florin, he advised her to set it on the number 3. Round went the wheel, and in twenty seconds the ball tumbled into compartment 3 sure enough. At the next turn she asked his advice, and was told to try number 24. No sooner said than done, and 24 came up in due course, whereby Mdlle L. C. won 140 odd gulden in two coups, the amount risked by her being exactly four florins. Like a wise girl, she walked off with her booty, and played no more that day at Roulette. A few minutes later I saw an Englishman go through the performance of losing four thousand francs by experimentalizing on single numbers. Twenty times running did he set ten louis-d'ors on a number (varying the number at each stake), and not one of his selection proved successful. At the "Thirty and Forty" I saw an eminent diplomatist win sixty thousand francs with scarcely an intermission of failure; he played all over the table, pushing his rouleaux backwards and forwards, from black to red, without any appearance of system that I could detect, and the cards seemed to follow his inspiration. It was a great battle; as usual, three or four smaller fish followed in his wake, till they lost courage and set against him, much to their discomfiture and the advantage of the