In order to encode/decode secret messages, clear-text/cypher-text was entered into a keyboard. Each letter was transformed in-turn according to the state of the internal switching mechanism. The active elements of this mechanism being a sequence of rotors, a reflector and a plug board, each of which were independantly configurable.

When an electrical current was passed through the rotors, they each applied a particular letter-substitution, according to their rotational position and internal wiring. Having passed through the sequence of rotors, the signal was then mapped by a 'reflector', before being passed back through the rotors in the reverse direction. There are a number of points worth noting:

The machine could encode/decode messages made up of the letter symbols A-Z.
Numbers were spelt out in clear-text, and spaces were represented by the letter 'X'.
The encoding/decoding processes were mutually symmetric.
It was not possible to encode a letter as itself.

The Strength of Enigma

The strength of the approach was due to the fact that as each letter was processed, the position of the rotor mechanism moved, so that a different set of letter mappings were applied to each character in the sequence. It is this feature which renders simple letter-frequency analysis, which can be used to break static substitution (or caesar) cyphers, ineffective.

As each letter was encoded, the end rotor rotated by one notch, and at a preset position, this 'knocked' the adjacent rotor on by one position, and so on. This process included a couple of 'quirks':

Some rotors advanced twice per rotation of their adjacent rotor.
When processing a letter, the mechanism was advanced before that signal was passed through the machine, so the mappings that were applied were not those for the displayed rotor positions, but for the next ones.

So, without a detailed knowledge of the machine's construction and internal state, each letter mapping is in effect randomised with respect to the others, and no useful long-term structural patterns are apparent.

The Original Codebreakers

Most of the initial work in attacking this code was done by the Poles, who, using various message intercepts managed to deduce some details of the internal rotor wirings, and managed to break an early version of Enigma. They then passed their work onto the British who, as is now well known, built on these successes at Bletchley Park, using electro-mechanical computing devices.

It is perhaps worth noting that, even the ingenious techniques that were developed to break this code might not have actually succeeded, if it had not been for weaknesses in the way that it was used by the Germans. This made breaking Enigma considerably easier, by substantially reduced the number of machine configurations that needed to be tested.

Note: The photographs used here are reproduced with the permission of Tony Sale, who is their copyright owner.