The Capital Q.

Take a number of coins, say from five-and-twenty to thirty, and arrange them in the form of the letter Q, making the “tail” consist of some six or seven coins. Then invite some person (during your absence from the room) to count any number he pleases, beginning at the tip of the tail and traveling up the left side of the circle, touching each coin as he does so; then to work back again from the coin at which he stops (calling such coin one), this time, however, not returning down the tail, but continuing round the opposite side of the circle to the same number. During this process you retire, but on your return you indicate with unerring accuracy the coin at which he left off. In order to show (apparently) that the trick does not depend on any arithmetical principle, you reconstruct the Q, or invite the spectators to do so, with a different number of coins, but the result is the same.

The solution lies in the fact that the coin at which the spectator ends will necessarily be at the same distance from the root of the tail as there are coins in the tail itself. Thus, suppose that there are five coins in the tail, and that the spectator makes up his mind to count eleven. He commences from the tip of the tail, and counts up the left side of the circle. This brings him to the sixth coin beyond the tail. He then retrogrades, and calling that coin “one,” counts eleven in the opposite direction. This necessarily brings him to the fifth coin from the tail on the opposite side, being the length of the tail over and above those coins which are common to both processes. If he chooses ten, twelve, or any other number, he will still, in counting back again, end at the same point.

The rearrangement of the coins, which is apparently only intended to make the trick more surprising, is really designed, by altering the length of the tail, to shift the position of the terminating coin. If the trick were performed two or three times in succession with the same number of coins in the tail, the spectators could hardly fail to observe that the same final coin was always indicated, and thereby to gain a clue to the secret. The number of coins in the circle itself is quite immaterial.