WhereIsAtlas · Technology

Where Is the Enigma Machine Now?

The company that built it went out of business in 1935 and its successor closed in 1945. The machine did not — it is in museums, with catalogue numbers, and it still sells at auction for six figures.

Enigma was a family of German rotor cipher machines, patented in 1918 and built from 1923 by a Berlin company that went out of business in 1935. Surviving machines are museum objects — Bletchley Park holds eight with their model and serial numbers, the Deutsches Museum holds two M4s, and a 1944 Olympia-built M4 sold at Bonhams for £305,200 in September 2025. The more consequential lineage is the one it provoked: Polish mathematicians reconstructed the machine in the early 1930s, handed their work to Britain and France in July 1939, and the machines built to attack it led to Colossus and to modern computing. Four other German firms built Enigma under licence, and their manufacturer codes are stamped on the machines that survive.

The question has two answers

"Where is the Enigma machine now?" sounds like one question. It is two, and they have different answers.

The first is about the object. There is no Enigma machine in service, because there is no organisation that uses one and no company that makes one. The manufacturer was liquidated in 1935 and its successor stopped trading in May 1945 [2][3]. What survives is a limited number of physical machines, mostly in museums, with catalogue numbers, serial numbers and — on the German ones — a three-letter code identifying which factory built that particular unit.

The second question is about consequence, and it is the more interesting one. The machine's place in history is not what it protected. It is what it forced other people to build. Reconstructing and then attacking Enigma produced the first electromechanical key-search machines, then the electronic Colossus, and the people who built those went on to build the first computers. The lineage that matters runs out of Enigma and into computing.

What the machine was

Enigma is not one machine. It is a family of models sharing a design, and the word covers commercial machines sold in the 1920s and 1930s as well as the military versions of the Second World War [5].

The design is a substitution cipher whose substitution changes with every keystroke. Pressing a key completes a circuit. Current leaves the keyboard, passes through a plugboard, then through three rotors from right to left, reaches a reflector, returns through the same rotors in the opposite direction, passes back through the plugboard, and lights a lamp. There is no printer — the machine's only output is which lamp lights.

Each rotor is a disc with 26 contacts on each face, internally wired so that each contact on one side connects to a different contact on the other. That is a fixed substitution alphabet, but the rotor turns, so which wire a given letter enters changes with its position. Three rotors in series compose three substitutions, and because the right rotor turns before every letter, the composition differs for every letter of a message.

Two design facts matter later. The reflector pairs up the 26 contacts, which makes the machine its own inverse — the same settings both encipher and decipher — and guarantees that a letter can never be enciphered as itself. That second property is not a rounding error; it is the crack that let the machine be attacked mathematically, because it means the substitution has a structure that pure chance would not produce.

The wartime variations matter too. The standard German military machine, Enigma I, took three rotors chosen from five. The navy's four-rotor M4, introduced without warning on 1 February 1942, halved the width of the reflector and used the freed space for a fourth, thinner rotor, with three additional rotors to choose from [7]. Bletchley Park's reading of U-boat traffic, codenamed Shark, went dark for nine months [4].

The company that built it

The patent was filed in Germany on 23 February 1918 and granted on 8 July 1925 [1]. It is almost always described as "Scherbius's patent", which is right about the inventor and wrong about the applicant. The patent was applied for by Gewerkschaft Securitas of Berlin, the body Arthur Scherbius traded under, and the US family member records Scherbius as the inventor assigning to that firm [1]. If you go looking for the patent under his name, you will not find it.

Chiffriermaschinen Aktiengesellschaft was founded on 9 July 1923 at Steglitzerstraße 2 in Berlin to build and sell the machine [2]. Note the three dates that circulate for this story and get conflated: 1918 is the patent application, 1920 is the Scherbius & Ritter partnership that preceded the company, and 1923 is the company [2][3]. The firm was also licensing abroad as early as December 1923, in negotiations with a Swiss manufacturer — records of the company's licensing activity, as distinct from its sales, survive in the patent and corporate record [30].

Commercially it was a disappointment. The machines were sold to businesses and governments, but the largest German users turned out to be government departments — the Abwehr, the SD and the Reichsbahn — and most commercial machines ended up with German authorities rather than foreign customers [3]. The best-documented early foreign sale is Swedish: two Enigma B machines ordered in January 1925 and delivered that April, after the Swedish General Staff borrowed two machines to evaluate them [9][31]. Switzerland is the best-documented larger customer, with 265 machines in use by July 1942 [8]. Spain received machines through German government and military channels during the Civil War, and one surviving Spanish example — a commercial model K dated 1931 — is now catalogued in the Spanish Army Museum's collection [33][34].

The company did not live to see its product become famous. It was liquidated on 31 July 1935 and its business was taken over by Heimsoeth & Rinke [2][6].

Who actually built them

This is the part that resembles the brand-and-factory questions elsewhere in this registry, and it is unusually well documented, because the German army stamped a manufacturer code on every machine it accepted.

Heimsoeth & Rinke did not do all the manufacturing. Records of the Heereswaffenamt's manufacturing-code list identify four other German firms building the machine under licence [10]:

| Code | Firm | Place | | --- | --- | --- | | aye | Olympia Büromaschinenwerke AG | Erfurt | | bac | Ertel-Werk | Munich | | gvx | Konski & Krüger | Berlin | | jla | Heimsoeth & Rinke | Berlin | | jmz | Atlas Werke AG | Bremen |

Those codes are still legible on surviving machines, which is how a museum object can be traced to the factory that made it. The clearest example is an Enigma M4 held by the Museumsstiftung Post und Telekommunikation in Germany, whose serial reads M 18311 / Na 220 900 / 18311/aye/44 — the `aye` identifying Olympia in Erfurt, the `44` the year [24].

What is not documented is foreign licensed manufacture. Other countries received machines — that is recorded in a 1943 German listing that names Italy, Romania, Hungary, Slovakia and Bulgaria as holding plugboard machines, with deliveries to Argentina, Japan, Hungary, Chile, Sweden, Holland and Czechoslovakia [3]. But no source found for this entry shows any foreign firm building Enigma under licence. The distinction is worth keeping: buying the machine is not the same as being allowed to make it.

The Poles got there first

The most consequential work on Enigma was not done in Britain, and getting this right is a matter of record rather than sentiment.

The Polish Cipher Bureau — a section of military intelligence — ran a cryptology course for mathematics students in Poznań from January to May 1929. Marian Rejewski, Jerzy Różycki and Henryk Zygalski were among the outstanding participants; Rejewski and Zygalski were employed at the Bureau's Poznań outpost in autumn 1929, and Zygalski was transferred to the German-ciphers section, Referat BS-4, in September 1932 [11].

Rejewski's achievement was mathematical rather than mechanical. Working from intercepted traffic, he reconstructed the machine's internal wiring without ever having a machine to take apart — the reconstructed design, and the methods built on it, are the substance of his own published account, "An application of the theory of permutations in breaking the Enigma cipher" [13]. The National Museum of Computing states plainly that the Poles deduced the wiring of the five wheels and that their cryptologic bomb, the bomba kryptologiczna, "was nothing like the Turing-Welchman Bombe" [14].

That last sentence is the one to hold on to. The Polish bomba of 1938 and the British Bombe of 1940 share a name root and an idea — exploiting a known or guessed piece of plaintext to test rotor settings mechanically — and they are different machines built in different countries [12][14]. Conflating them is the most common error in retellings of this story.

The handoff

In July 1939, with war weeks away, the Poles disclosed what they had. Polish, French and British cryptanalysts met outside Warsaw on 25 and 26 July 1939; Britain's Alastair Denniston and Dilly Knox were there, and the Polish work was handed over [12].

A caution on the date, because it is repeated inconsistently. GCHQ's own material contradicts itself: its article on the Pyry Forest meeting gives 25–26 July 1939, while a 2014 speech by a former GCHQ director on the same site gives 15 and 16 July [12]. The commonly repeated "25–27 July" does not appear in either. The National Archives holds the file of papers relating to these pre-war meetings between French and Polish cryptanalysts and GC&CS, covering 1938 to 1943, so the sequence is documented even where the day is not [20].

What the Poles handed over was not a machine. It was a reconstruction, a set of methods, and several years' head start. Britain's advantage from 1940 onward was industrial capacity — the ability to build key-testing machines by the hundred and to staff the work at scale — rather than a mathematical breakthrough that had not already happened in Warsaw.

Bletchley Park

The attack was industrialised at Bletchley Park, the wartime home of GC&CS. Alan Turing wrote the foundational technical account, catalogued by The National Archives as HW 25/3, "Mathematical theory of ENIGMA machine by Alan Mathison Turing", known at the time as "Prof's book", with covering dates of 1939 to 1942 [17]. Note the dates: the document is not a single 1940 treatise, which is how it is usually described.

The Bombe was the machine that made the attack scale. It tested rotor settings against a guessed piece of plaintext, and Gordon Welchman's contribution — the diagonal board, published in his account The Hut Six Story — exploited the reciprocal property of the plugboard to eliminate far more wrong settings per test than the original design could [15][22]. The first Bombe became operational in March 1940 [15]; the Archive's Bombe Register for 1940 to 1951 is held as HW 25/19, and the catalogue describes the Bombe as "developed by Alan Turing and others" [18].

Colossus was not part of this. It is routinely lumped in with Enigma, and it is a different story. Colossus was built to attack the Lorenz cipher — the German high-command teleprinter traffic the British called Tunny — and The National Museum of Computing states it "had a single purpose: to help decipher the Lorenz-encrypted (Tunny) messages" [16]. The National Archives catalogues Colossus alongside Tunny, and holds photographs of Enigma, Tunny and Colossus as three distinct subjects [19]. Breaking Enigma and breaking Lorenz were separate problems with separate machines.

One note on the sources for this section, since it matters for reuse. The material cited from The National Archives is Crown copyright released under the Open Government Licence, which is why the catalogue references are given in full rather than paraphrased — the records themselves are the citation, and they are free to read [21].

Where the machines are now

Surviving Enigma machines are museum objects, and several institutions publish enough detail to identify individual units.

Bletchley Park holds the most comprehensive display and names eight machines with both model and serial number [4]: an Enigma D (A1214) delivered to the Italian Navy in 1932 and captured in northern Italy in 1945; an Enigma K (K289) from a batch sent for the Nationalist side in the Spanish Civil War, used at a military headquarters in the Canary Islands; two Enigma I machines (A02192 and A16992, the latter heavily modified by the codebreakers); an Enigma M4 (M5846); an Enigma T (T244); and two Enigma G machines (G312, an Abwehr machine, and G110, from a 1931 shipment to the Hungarian armed forces).

The Deutsches Museum in Munich catalogues two M4 machines: inv. 2017-398, dated 1942 with the inscription "ENIGMA" and stamped 10132, on display in the exhibition Bild Schrift Codes; and inv. 1984-584, built around 1944 and inscribed M 7972, which was CT-scanned in 2023 as a research dataset [23].

The Museumsstiftung Post und Telekommunikation holds the M4 whose serial ties it to licensed production — the Olympia-built machine described above [24].

The Polish Army Museum in Warsaw holds several, and its own account is unusually specific: a three-rotor commercial machine transferred by the Ministry of the Interior in March 1976; a four-rotor military machine found dismantled by a 17-year-old, Zbigniew Kosecki, in a pigsty near Opoczno and bought in 1976 for 4,000 złoty; and a machine given in 2000 by the Duke of York to Prime Minister Jerzy Buzek, recovered from a German U-boat [25].

Others include the Australian War Memorial, which displays the common three-rotor Wehrmacht and Luftwaffe model [26], and the Powerhouse in Sydney, which catalogues a 1943 Wehrmacht machine as object 94/90/1 [27].

One post-war destination is easy to miss. The Israeli Signal Corps museum records that the IDF bought 50 Enigma machines in the early 1950s and converted them to Hebrew [32] — a machine designed to protect German military traffic, repurposed by a new state's army less than a decade after the war ended.

What they sell for now

Enigma machines come to auction, and the prices are well documented when the auction house publishes them.

In September 2025 Bonhams sold a 1944 four-rotor M4 built by Olympia Büromaschinenwerke for £305,200 including premium [29]. In November 2025 Christie's in Paris sold an M4 dated November 1941, serial M5426, for €482,600 against an estimate of €200,000–300,000 [28]. The Christie's price rests on the auction house's own press release; its lot page confirms the serial and the rotor serials but does not display the result [28].

Two figures in circulation are wrong and should not be repeated. A widely shared "$547,500" is not an Enigma price at all — it is the total turnover of an entire auction. And a "$269,000" figure attributed to Bonhams in 2015 traces to a pre-sale press release with an estimate of £30,000–50,000. Where a number cannot be tied to an auction house's own published result, this entry does not use it.

What the record does not settle

Three things are genuinely unresolved, and they are worth stating rather than smoothing over, because the popular accounts present all three as settled.

When the German military adopted the machine. Bletchley Park says German armed forces adopted Enigma from 1926 [4]. Crypto Museum gives 1926 on one page, a first batch of about 600 machines delivered on 10 December 1927 on another, and "put into service on 1 June 1930" on a third [5][6]. These are not rounding differences.

When the plugboard appeared. It is commonly dated to 1930, which is wrong. The documented sequence is a single-ended patch panel in early 1927 and the final double-ended Steckerbrett design in 1928, with the finished Enigma I entering service later [5][6]. 1930 is the date of the finished version or of service entry, not of the plugboard's invention.

How many were made. The figure usually quoted — 100,000 or more — is not supported. The best-sourced number is 35,616 plugboard machines of all types delivered to the Wehrmacht, from research built on captured German technical-intelligence documents [3]. That figure comes from a single researcher's work and has not been independently restated, so it should be attributed rather than repeated as settled fact.

The short version

A machine patented in 1918 by an engineer whose name is on it but who did not apply for it, built from 1923 by a Berlin company that failed commercially, adopted by an army, and manufactured under licence by four other German firms whose codes are stamped on the survivors. Reconstructed from intercepted traffic by a Polish mathematician who never saw the inside of one, handed to Britain and France in July 1939, and attacked at scale by machines built for the purpose — machines that led, by a route that runs through Colossus, into computing.

The company died in 1935. The machine is in museums, and the consequences are in every computer.

Operate one. WhereIsAtlas has a working Enigma I model at its own address on this site — /enigma — with rotors I to V, both reflectors, ring settings, the plugboard and the real middle-rotor double-step. It is a demonstration of the mechanism described here, not encryption, and it never sends anything anywhere.

Entries in this story

Sources

  1. DE 416 219 C — Chiffrierapparat, filed 23 February 1918, granted 8 July 1925; applicant Gewerkschaft Securitas — Google Patents (2026-09-21) ↩
  2. Chiffriermaschinen Aktiengesellschaft — Crypto Museum (2026-09-21) ↩
  3. Enigma development and production history — Crypto Cellar Research (Frode Weierud) (2026-09-21) ↩
  4. Enigmas of Bletchley Park — models and serial numbers of the machines held — Bletchley Park Trust (2026-09-21) ↩
  5. Enigma I — Crypto Museum (2026-09-21) ↩
  6. Enigma history — the plugboard's development — Crypto Museum (2026-09-21) ↩
  7. Enigma M4 — Crypto Museum (2026-09-21) ↩
  8. Enigma K in Switzerland — Crypto Museum (2026-09-21) ↩
  9. Enigma A — the early commercial models — Crypto Museum (2026-09-21) ↩
  10. Enigma manufacturing codes (Fertigungskennzeichen) — Crypto Cellar Research (Frode Weierud) (2026-09-21) ↩
  11. Henryk Zygalski (1908-1978) — the Poznań cryptology course and Referat BS-4 — Instytut Pamięci Narodowej (Polish Institute of National Remembrance) (2026-09-21) ↩
  12. The Pyry Forest meeting — GCHQ (2026-09-21) ↩
  13. An application of the theory of permutations in breaking the Enigma cipher — Rejewski, Applicationes Mathematicae 16 (1980), 543-559 — Instytut Matematyczny PAN (IMPAN) (2026-09-21) ↩
  14. The Polish connection — The National Museum of Computing (2026-09-21) ↩
  15. Enter Turing and Welchman — The National Museum of Computing (2026-09-21) ↩
  16. Colossus — The National Museum of Computing (2026-09-21) ↩
  17. HW 25/3 — "Mathematical theory of ENIGMA machine by Alan Mathison Turing", covering dates 1939-1942 — The National Archives (UK) (2026-09-21) ↩
  18. HW 25/29 — the Bombe — The National Archives (UK) (2026-09-21) ↩
  19. HW 25/34 — Colossus and Tunny — The National Archives (UK) (2026-09-21) ↩
  20. HW 25/12 — pre-war meetings between French, Polish and British cryptanalysts — The National Archives (UK) (2026-09-21) ↩
  21. Crown copyright and the Open Government Licence — The National Archives (UK) (2026-09-21) ↩
  22. The Hut Six Story: Breaking the Enigma Codes — Gordon Welchman — Open Library (2026-09-21) ↩
  23. Chiffriermaschine Enigma M4, Inventarnr. 2017-398 — Deutsches Museum (2026-09-21) ↩
  24. Enigma M4, Inventarnr. 4.0.23036, serial M 18311 / aye / 44 — Museumsstiftung Post und Telekommunikation (2026-09-21) ↩
  25. Niemiecka maszyna szyfrująca Enigma w Muzeum Wojska Polskiego — Muzeum Wojska Polskiego (2026-09-21) ↩
  26. Enigma — Australian War Memorial (2026-09-21) ↩
  27. Enigma cipher machine, object 94/90/1 — Powerhouse (Sydney) (2026-09-21) ↩
  28. The Exceptional Sale, Christie's Paris, 19 November 2025 — Enigma M4 serial M5426 — Christie's (2026-09-21) ↩
  29. Science, Technology & Natural History, Bonhams, September 2025 — an Olympia-built M4 at £305,200 — Bonhams (2026-09-21) ↩
  30. Scherbius and the Enigma: political, economic and military conditions — Claus Taaks, HistoCrypt 2023 — Linköping University Electronic Press (2026-09-21) ↩
  31. The first classical Enigmas: Swedish views on Enigma development 1924-1930 — Anders Wik, HistoCrypt 2018 — Linköping University Electronic Press (2026-09-21) ↩
  32. Cipher Machine TG-356 (Enigma) — the machine bought by the IDF in the 1950s — Israeli Signal Corps Museum (2026-09-21) ↩
  33. Enigma in Spain — Crypto Cellar Research (Frode Weierud) (2026-09-21) ↩
  34. Máquina de cifrar alemana marca Enigma, tipo comercial, modelo K, 1931, inv. MUE-202663 — Museo del Ejército (Spanish Ministry of Defence) (2026-09-21) ↩