IKA Cipher Machine
Cipher traffic can reveal a system while leaving its physical mechanism uncertain.
Historical naval cipher reconstructed from intercepted traffic; the physical IKA mechanism is not established by the cited 2021 study.
Why This Matters
A working reconstruction can reproduce a cipher’s output without reproducing the original machine. IKA makes that distinction concrete: historical reports describe its cryptographic behavior, while familiar photographs belong to other devices. An artifact’s caption is part of the evidence.
What the Traffic Reveals
Chris Christensen’s 2021 study draws primarily on the US Navy’s April 1946 report RIP 28A. IKA’s 49-symbol kana system separates a seven-symbol group with fixed substitution from a 42-symbol group with a monthly sequence and staggered movement. Its documented output behavior allows cryptographic reconstruction.
The paper lists US intercepts from late 1931 or early 1932 to July 1933. Those are intercept dates, not a known invention date. It reports that US Navy codebreakers never captured or saw an IKA machine.
Four Kinds of Evidence
| Evidence | Supports | Limit |
|---|---|---|
| Intercepted IKA traffic and reports | Reconstructed substitutions and stepping behavior | Does not identify the physical input, output or stepping mechanism |
| Captured M-1 / ORANGE hardware | The successor machine’s physical design; Figure 2 shows M-1 | A surviving M-1 is not an authenticated IKA |
| Holtwick’s American M-1 analogue | An American implementation; shown in Figure 3 | The study rejects its occasional identification as IKA |
| Proposed IKA mechanisms | Possible realizations of similar cipher behavior | Kryha-like disks or a Damm half-rotor remain hypotheses |
Scroll the table sideways to read all three columns.
No artifact photographs are reproduced here. Consult the paper for the photographs and their attributed captions.
What a Replay Would Leave Open
Two mechanisms can implement the same substitutions and stepping sequence. Matching every available ciphertext would test that mathematical behavior; establishing which mechanism the original makers used would require additional physical or documentary evidence. This exhibit offers no machine simulator.
The 2021 study also reports that records of the original cryptanalytic attack were unavailable. A modern reconstruction should therefore identify its own method and assumptions, rather than present them as the historical attack.
Try the Charles V evidence exercise: a partial key can pass a check while still leaving an unknown symbol. Follow the Portuguese letters of 1649 to see why reading, translation and historical consequences need separate evidence.
Sources and Reading
- Chris Christensen, “The Imperial Japanese Navy IKA Cipher Machine”, HistoCrypt 2021, DOI 10.3384/ecp183155. See §2–4 for the reconstructed cipher, §5–7 for hardware and analogue attribution, and Appendix A for intercept dates and alphabet splits.
- Underlying reports cited by Christensen: Chief of Naval Operations, The M-1 Machine System: RIP-28A (April 1946), NARA RG 38, Box 16; Jack S. Holtwick Jr., History of the Naval Security Group to World War II: SRH 355 (1971), especially Appendix VIII, 90F–90G. The museum has not independently inspected these archival reports.
The evidence limits above are those of the cited 2021 study. They are not a claim to have checked every surviving collection or later archival discovery.