Runs, intermittencies, and other tendencies of chance at certain tables could be carefully noted; occasionally there would be no play at all at one table, the whole crowd staking on a run at another; as the room was small, anything of the sort soon reached the ears of every one. Play as a rule was high, and the players, for the most part, were well used to gambling. The results to the bank were most disastrous. On a certain evening it lost more than had ever before been lost in one day by the Casino, and at the end of the year the accounts of the "Cercle Privé" proved anything but an agreeable study for the officials supervising the finances of the great gambling monopoly. The next year it was closed, and there has since been no inclination on the part of the authorities to repeat what was to them a very unprofitable speculation. Amongst various causes which in this instance operated to the detriment of the bank was the difficulty, generally amounting to impossibility, of players obtaining a further supply of money when what they had in their pockets had run out. At such a late hour, when the Bank was closed and the _caisse_ of most hotels shut up, no matter how rich a man might be, he could not obtain any considerable amount of cash. Consequently, should he lose what he had brought with him, he was reduced to playing with such modest sums as could be borrowed from friends, who naturally could not be expected to make any substantial advance, as any moment they themselves might be in a similar predicament. The bank, on the other hand, was equipped with ample funds, and its loss--unlike those of the players, which, after a certain point, were limited by necessity--often extended into a very large figure; consequently, when it was in good luck, it only won a comparatively moderate amount, and when in bad lost very heavily. Another reason for the ill-success of the bank was that the policy pursued in the large rooms downstairs had in the case of the "Cercle Privé" been exactly reversed. In the former there have always been many more roulette tables than tables devoted to trente-et-quarante--upstairs there was only one roulette table as a counter-attraction to the three devoted to the rival game. Trente-et-quarante is mathematically one of the most favourable of games at which a gambler can play, the percentage against him produced by the _refait_ being only 1·28 per cent. Roulette, on the other hand, is, owing to the zero, highly advantageous to the banker. The bank's percentage on all-round play at the tables is more than one-seventy-fourth of all the figures staked; the actual winnings of the bank being about one-sixtieth part of all the money actually placed on the board. At the present time the bank's winnings (gross) are, roughly, £1,200,000 per annum.