A pack of cards is handed out to be shuffled. When it is returned the performer learns what the top card is by simply bending the left hand bottom corner, so that he may see the index. Laying the pack on the table, face down, he asks some one to divide it in two. He informs his audience that by looking at a card in one packet he is enabled to tell which card is on top of the other packet.
In proof of this he lifts a few cards from the packet which was the lower part of the pack, and looking at one card he names the top card of the other packet. As he knows that card, this is an easy matter. Looking at the cards of the lower packet is merely a ruse to learn what card is on top, for while the audience imagine he is looking at one of the cards when he opens the packet he is really shifting the top card so that he may see its index, as shown in Fig. 47. When the performer has named the top card of the upper packet and it is found to be correct, he deliberately places the former lower packet on top. Then he shuffles the pack without disturbing the top card, and proceeds as at first. In this way he may continue indefinitely, without fear of discovery.
He is careful to give his audience the impression that when he lifts at hazard the cards from the lower packet, as already shown, that the particular card he sees enables him to tell the top card of the upper pack. To strengthen this impression, when he looks at the card of the lower pack he says: "Ah! the five of diamonds—" or whatever it happens to be—"naturally, the top card of the other packet must be the Jack of Clubs, just as naturally as day follows night." It is wonderful how this little subterfuge tends to mislead even intelligent people; most of them will imagine that the trick is the result of some mathematical formula worked out by the performer. Not for a moment must any one be led to suspect that the performer's sole object in handling the lower packet is to get sight of its top card.