In his autobiography, Silent Years[1], John F. Byrne, a lifelong friend of James Joyce, devoted the last chapter to Chaocipher which he had invented in 1918. Byrne described his attempts starting in 1920 to interest the State, War, and Navy Departments in his indecipherable cipher and his frustration with the disinterest shown by William F. Friedman and other cryptanalytic experts after he had demonstrated his machine.

In a letter to Friedman, February 17, 1957, who had written earlier about “why you put in that book the completely extraneous matter of ciphers when you had so much of interest to tell of Joyce,” Byrne admitted that “my main reason for writing the book was to be on record in the matter of my ‘Chaocipher’.” Byrne was still so convinced of its indecipherability that he included an enciphered message in his book and offered a $5,000 reward to anyone who solved it within three months after the book was published. He specifically challenged Bell Laboratories, where he had also demonstrated his machine, all “cryptanalytic and cryptographic” experts, and the American Cryptogram Association and its local affiliate, the New York Cipher Society, to solve the cipher. The reward was never claimed.

Since then, however, many attempts have been made to solve the Chaocipher. (For an excellent analysis by G. E. Mellen see [4].)

Over the years many hypotheses about the cipher have been suggested and most of them were demolished in Mellen’s article, e.g., it is a crude rotor system, contains an element of transposition, autokey cipher, Vernam tape system, and a form of polyalphabetic cipher with a random nonrepeating key.

An overlooked, possibly valuable clue is in The Editor’s Notebook of the American Cryptogram Association 1952-1956. This is a detailed chronological diary kept by Henry E. Langen who was editor of the ACA’s magazine, The Cryptogram, during that period. The diary contains a letter from Byrne, December 10, 1953, in answer to Langen’s request for a demonstration of Chaocipher. Byrne said he would be delighted to demonstrate his apparatus and after many postponements, Byrne, accompanied by his wife, visited Langen in New Jersey on May 25, 1954. For two hours he discussed his book and Chaocipher but he did not bring the cipher machine “explaining that it was too heavy and cumbersome.” He did bring blueprints which Langen did not understand. According to Langen, “He did explain that the machine is made up somewhat like a typewriter with two revolving disks with the alphabets arranged along the periphery in a complete disorder.” Langen commented parenthetically that “With only two disks used; I am a bit confused as to how this can result in such utter chaotification of the plaintext message.”

In 1973, Louis Kruh, seeking an interesting program for the 1974 annual convention of the ACA which he had volunteered to organize, thought of the unsolved Chaocipher. If he could obtain information on Chaocipher from Byrne’s family it would make a fascinating program. Thus began a search for Byrne’s heirs. After several months of research a letter about his hunt which appeared in the Irish Echo, resulted in a letter from John Byrne, son of John F, Byrne, with his telephone number. When contacted, John Byrne said he had helped his father with the cipher and was totally familiar with it. He said that perfect accuracy was essential when enciphering a message because one error would distort the rest of the cipher. But, unfortunately, he did not want to reveal how Chaocipher worked.

Kruh, not easily discouraged, continued to call or write to Byrne periodically for the next 15 years. In 1978, Byrne said his mother had died and that he was sorting out his father’s papers. He also expressed concern that if Chaocipher had commercial application it shouldn’t be taken from the family which was why he was reluctant to explain how it worked. In a call near the end of 1989, Byrne agreed that he had to make a decision on what to do with Chaocipher, i.e., determine whether it had commercial value and proceed accordingly or reveal its methodology in an article. Kruh suggested they meet to discuss how Byrne might make that determination and that to assist he would ask Dr. Cipher A. Deavours to attend the meeting. He explained that Deavours was a professor of computer science, co-author with Kruh of Machine Cryptography and Modern Cryptanalysis [2], an accomplished cryptanalyst, and an internationally known consultant in computer security matters.