GREENBERG | 93
sender proceeded to disguise ARQ before transmitting it. To do so, he chose another
trigram from the K-book, YVT say. He then wrote down Y V T
A R Q
and filled in the two blanks _ with any two letters he cared to choose, giving (say):W Y V T
A R Q N
Next, he consulted the day’s bigram table and replaced the vertical pairs of letters, WA,
YR, VQ, TN, by their equivalents given in the table—say, IJ, TV, US, YX. Finally, he
placed IJTV USYX at the beginning of the encrypted message (see the previous exam-
ple of an intercepted message). IJTV USYX was called the message’s ‘indicator’. For
good measure, the operator added the indicator once again at the end of the encrypted
message.
3 The message’s official recipient looked up IJ, TV, US, YX in his copy of the bigram table
and replaced the letters by their equivalents (WA, YR, VQ, TN), giving:W Y V T
A R Q N
He then set his wheels to the ground setting (JNY in this example) and decrypted ARQ
to get LVN, the message setting. Finally, he set his wheels to LVN and decrypted the
message by (as usual) typing the message’s ciphertext at the keyboard; the German
plaintext would light up, letter by letter, at the lampboard.Technicalities: reciprocity and wheel ‘turnovers’
The Enigma machine was designed to be ‘reciprocal’. This means that if (say) the machine
encrypts O as Q, then typing Q at the same position of the wheels produces O at the lampboard
(see Fig. 12.5). It is because Enigma was reciprocal that the two processes of encryption and
decryption were essentially the same. Typing the plaintext produces the ciphertext, and typing
the ciphertext at the same machine settings produces the plaintext again.
The reciprocal nature of the machine was a consequence of the design of the reflector and
the plugboard. There was, however, a cost to the way the reflector worked. Because of the reflec-
tor’s design, it was impossible for a letter to be encrypted as itself—a crucial weakness of the
machine. For a letter to encrypt as itself, the electrical current would have to pass along a wire
in both directions at the same time, which is a physical impossibility. It seems that the Germans
were prepared to accept this weakness, presumably because they viewed it as a tolerable price
to pay for a reciprocal design. A machine whose processes of encryption and decryption were
identical was so much simpler to operate. However, the machine’s designers completely failed
to appreciate how useful this weakness would be to the codebreakers (see Chapter 12).
The function of the wheel ‘turnovers’ was to ensure that the machine’s wiring varied with
every key press, or in other words to make sure that at least one wheel moved every time a letter
was typed at the keyboard. The right-hand wheel always turned over one step each time a key
was pressed at the keyboard, irrespective of the motion of the other wheels. The middle wheel