[Next] [Up/Previous] [Index]

Cliques on the Rods

Before the plugboard was added to the Enigma, the fact that the fast rotor was on the side opposite the reflecting rotor meant that it was vulnerable to an isomorph attack. But the isomorph attack had to be modified from that used on a Hebern machine, in the obvious way.

Both the probable plaintext and the ciphertext had to be put through the fast rotor, and it was the two results that had to match, that is, be monalphabetic encipherments of one another, having repeated letters in the same places.

Since a rotor can be in 26 different positions, changing both the input and output letters, one needs 26 different strips. Each strip shows, for one particular input letter, the possible output letters for each position of the rotor.

Thus, if a rotor is wired as follows:

```A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
E K M F L G D Q V Z N T O W Y H X U S P A I B R C J
```

the strip for the letter A would contain the sequence

```E J K C H...
```

and the strip for the letter B would contain the sequence

```K L D I...
```

since in the A position of the rotor, A becomes E, the A strip begins with E. When the B contact, that takes B to K in the A position, is moved back to the A position, then it will take A to the letter before K, or J. Similarly, when the C contact is moved two positions back, its destination, the M contact, is moved two places back to K. Of course, the strips might also be prepared for the rotor moving the other way, with the letters in reverse order.

The strips will have their sequences of letters repeated twice; one chooses the strips which correspond to plaintext and ciphertext, and puts them next to one another, but staggered, because the fast rotor is always moving one step for every letter. Then, the rows across, between the strips, show the possible encipherments of the plaintext or ciphertext letters for every starting position of the fast rotor.

Strips with the rotor outputs on them were aligned to produce the results, and thus the method was called "La Méthode des Bâtons" in French, and in Britain the strips were called rods, with the repeated letters that matched up called cliques.

The sequence of 26 letters that a rotor would produce from a given letter in different positions cannot be a scrambled alphabet, with no letter repeated, for reasons seen in the previous section on the interval method. In the colorful terminology of Bletchley Park, the duplicated letter on a rod resulting from two wires producing the same displacement was called a "beetle", and the result of two wires having exactly opposite displacements was called a "starfish", as recently revealed in papers by C. H. O'D. Alexander available on Frode Weierud's web page. These papers note, among other things, that the rods were sometimes used against Enigmas with plugboards as well, in cases where the indicators could not be interpreted. (Two of the papers, about something called JNA-20, appear to be about PURPLE or CORAL, even though they still use the term "Stecker" in reference to its plugboard.)

[Next] [Up/Previous] [Index]