THE HAWK
Over the past several years, I have developed a number of "impossible" card locations,
dubbed the Birds Of Prey. To date, I have held back from publishing any of these. Here is
one that I hope you like. [Still under wraps: the Vulture, the Falcon, and the Vicious
Hummingbird...]
The effect is as follows: while the performer's back is turned, two spectators each remove,
note, and replace cards. The deck is shuffled by a spectator, then returned to the performer.
The cards are dealt face down, until the performer stops "on impulse"  successfully
stopping on one of the selections. A spectator re-shuffles the pack, as the performer offers
to repeat the location. However, the performer suddenly opts for a more impressive finish:
he names the second selection. The pack may be examined, for It is ordinary, with nothing
added or taken away.
A combination of principles is involved. First, the deck is stacked in a known order.
Standard systems can be used, such as "Si Stebbins" or "Eight Kings". The stack need not
be cyclical.
Additionally, the pack must bear a one-way back design. Obviously, this should be
relatively subtle. Rather than use a picture-backed bridge pack, use a pattern back with a
natural one-way design, such as Tally-Ho Circle. Set the pack so that every other card
points in the opposite direction. [This alternating concept has been traced back as far as
Charles Jordan's "Premo Detection".]
To begin the routine, start by false shuffling the pack. Shuffles of a riffle variety should be
used, so as to set a precedent for a subsequent shuffle to be done by a spectator. Table the
pack, and turn your back on the proceedings.
A spectator is instructed to give the pack one or more complete cuts. Spectator A is told to
remove the top card of the pack, and note it. Spectator B takes the next card. A now
replaces his/her card on top of the pack; B does the same. [Note that these simple actions
re-position the two selections so that their one-way orientations are "out-of-synch" with
the rest of the pack. This marvelous idea comes from Theo. Annemann's "The Alternate
Detection", a location effect which, while simpler than the one under discussion, is
extremely strong.] The spectators again give the deck several complete cuts, and finally
one riffle shuffle.
The above actions having been completed, there should be no doubt in anyone's mind that
the chosen cards are quite irretrivably lost. At this point, turn around and take the pack.
Re-cap the sequence of events, explaining that you will avoid looking at the faces of the
cards. As this is said, casually spread the deck to note whether the top and bottom cards are
of the same directional orientation. If this is the case, do nothing. If not, cut the deck be-
tween two cards which point in the same direction to arrive at that situation.
Slowly deal the cards face down onto the table. You are seemingly trying to "feel the
vibrations"... Actually, you are observing the cards in pairs as the dealing takes place.
Each pair of cards will be heterogeneous [i.e., one pointing north, one pointing south]. At
some time during the deal, you will arrive at a pair with both cards pointing the same way.
Stop the deal between these two [i.e., when the first card of this pair is atop the tabled pile,
and the second is still on top of the talon].
Announce that the "vibrations" seem right - you feel certain that you've stopped on
Spectator B's selection. As this statement is made, gesture with the left hand, and under
cover of this movement, use an Overhand Peek to determine the identity of the top card
of the talon.
Spectator B is asked to name his/her card. If the noted card is named, turn up that card
showing that you did indeed stop on the correct card. If a different card is called, then you
know it is the top card of the tabled pile, and that card is turned up. In either case, you
construct the interpretation of the circumstance to "prove" you stopped at the precise
location necessary. [This either/or gambit is Dai Vernon's.]
Thus far we have made use of the One-Way Principle combined with the Gilbreath
Principle. It is here that the stacking preparation comes into play. You know the identity of
Spectator B's card, and thus can easily calculate A's selection  for it is the card that
was above B's card in the' original stack order.
As a way of further obscuring the methodology involved, at this point you hand the pack
to a spectator for further shuffling. In this fashion, the spectator does the dirty work for
you, completely destroying the evidence of the stack. [This is assuming you started with a
detectable rosary stack. In the case of a more obscure set-up this further shuffling is of
course not needed.]
Take the deck again, and begin to deal the cards face down as you did before. Suddenly,
as if seized by inspiration, announce that you will attempt to reveal the second selection
by much more difficult means. With proper histrionics, name A's selection.
The deck is clean; the routine is ended.!
ABACUS
I have long been intrigued by the Roy Walton "Palmist's Prophecy" plot ["Card Corner" in
Linking Ring, June 1973]. There have been very nice versions and variations by Swinford,
Hudson, Rogers, Mario, and others. It occured to me that the palindromic principle
exploited by Francis Haxton in his "Colour Forecast" [Hew Pentagram, December 1979],
just read at the time of this routine's devising, could be applied toward a modified version
of the Walton plot. While exploring the possibilities of such a marriage, the following
elaborate routine evolved. Along the way, the influence of Stewart James was clearly
present.
This routine will sound rather complicated upon first reading. I assure the reader that it is
not, in actual play. This write-up will be far more functional, however, if you follow along
with cards in hand.
A set-up is required. From the top of the pack: an ace; a king; six indifferent cards; the king
of hearts; one indifferent card; the next ten cards in numerical sequence (ace through ten);
ten indifferent cards; a king; a four (this card is crimped); a three; a two; fourteen indifferent
cards; a king; a two; a three; a four.
Begin by writing a prediction of the KH on a piece of paper. This is placed into a spectator's
keeping. Hand the pack to a spectator. He/she is instructed to think of any number from ten
to twenty. While your back is turned, the spectator silently deals that number of cards onto
the table, face down, and then covers the dealt pile with his/her hand. [We shall refer to this
pile as X.]
When you turn back around, take the balance of the pack. The cards are held in the left
hand, dealing position. The right hand grasps the pack from above, and the thumb contacts
the pack at the inner end. The right thumb releases three cards from the bottom of the pack,
and these cards are gripped with the left little finger.
The right hand maintains its hold on the pack. The left hand moves away to the left. The left
thumb draws away the top single card of the pack, and the left fingers draw away the three-
card block from the bottom. The right hand .stock is tabled, and the four cards in the left
hand are placed on top. These actions take but a moment, and appear to be no more than a
casual cut.
Immediately continue by cutting the pack into two piles,
cutting at the crimped card [i.e., the crimped card becomes the top card of one of the two piles]. We will refer to the pile
topped by the crimped card as Y, the other as Z.
There are now three portions of cards. One is beneath the spectator's hand, containing a number of cards determined by
(and known only to) the spectator. Explain that you will now have four cards taken from the other two portions, and that
these cards will perform a magical function. State that the spectator will determine which cards come from which pile.
Have the spectator point to either Y or Z. Remove the top card of the indicated pile. Again, the spectator points to one of
the piles/ and again the top card is removed. Repeat this twice more. [Note that the spectator has a genuinely free choice
in this  the only condition being that at least one card must be chosen from Z. This can be easily guaranteed by your
explanation prior to the choosing that the cards will be drawn from both piles, at the spectator's discretion.]
Turn up the four selected cards, and add their values together. The total will be the same as the spectator's secretly chosen
number. When this information has sunk in, turn the four cards face down and drop them on top of either Y or Z.
Instruct the spectator to raiae his/her hand from the counted pile. Pick up this packet. State that there is a faster way to
magically produce the chosen number. Now comes a move that dates back to Hofzinser: hold the packet in the right hand,
thumb on top, fingers below. The packet is now tossed to the table  but the thumb and fingers press inward, and retain
the top and bottom cards.
Turn over the two retained cards. They will represent the thought-of number [i.e., if the spectator thought of sixteen, they
will be an ace (one) and a six].
Say that you will take things a step further. Add the two cards' values together. [In the example just mentioned,
6+1=7.] Place the two cards face down onto either Y or Z. Pick up the spectator's pile (X), and count down to the
number just arrived at. The card at that position is dealt aside, and cards above it returned on top of X.
The card thus arrived at will be the king of hearts. Have the prediction message read.
For the climax, mention that you thought the king of hearts might be "lonely"  BO you provided company...
Turn over each of the three tabled piles. There will be a king on the face of each.
[This routine originally appeared in Apocalypse magazine.]