Enigma D
Rotor
Enigma
Tree
I →
Z →
A28 →
K →
← C
|
The machine has several improvements over the Enigma C.
The top lid was made more accessible by adding
a hinge at the rear and two retaining bolts at the front.
This made it easier to access the rotors and alter the internal settings.
There are four settable rotors that protrude the machine's top lid, which is why this machine is sometimes erroneously called a 4-rotor Enigma. In reality there are three driven cipher rotors (i.e. the rightmost three). They are mounted on a removable spindle, so that their order can be changed as part of the daily key settings. The leftmost rotor is the Umkehrwalze (UKW, reflector). It can be set to any of 26 positions, but is not driven by the cipher rotors. This is different from earlier machines, in which the UKW could be set to two or four positions, or could not be set at all. Enigma D is also the first Enigma machine on which the keyboard and the lamp panel are arranged in typewriter order (QWERTZ...) rather than the alphabet (ABC...). |
|
It seems likely that in 1926, the Enigma D was the core design on
which all future designs
would be (partly) based. The Enigma D itself
was short-lived and was succeeded a year later – in 1927 – by the
Enigma K (A27). Military Enigma I and Zählwerk Enigma A28
are both based on Enigma D.
The machine shown above has serial number A319 and was
(re)discovered in 2024
at an auction in Florida (USA). It is currently unknown who the
original user was. It is known however that its twin brother — serial
number A320 — was bought in 1926 or 1927 by the UK's codebreaking
establishment,
the Government Code & Cipher School (GC&CS, now: GCHQ).
It was subsequently scrutinised in 1927 by cryptanalyst
Hugh Foss, who wrote
a detailed report
and developed a method for breaking it [6].
Ten years later, in 1937 — during the Spanish Civil War (1936-1939) — it helped
codebreakers William Bodsworth and Dillwyn (Dilly) Knox
to break the Enigma traffic between the Spanish Navy of General Franco,
and the Italian Navy of dictator Mussolini [7].
The images below provide a quick overview of the features of Enigma D.
The image on the left shows the machine after opening the lid of the wooden
case. The device has three cipher rotors and a
settable reflector (Umkehrwalze, UKW).
To the right of the rotors is the
power selector.
It offers a choice
between HELL (bright) DKL (dunkel, dark), AUS (off) and SAMMLER (accumulator). 1
In the latter case, an external 4V power source should be applied to the two black
terminals at the right edge. 10 spare light bulbs
are stowed in a bracket inside
the upper edge of the case lid.
The image on the right shows the same machine with open lid. This exposes the
cipher rotors, the ETW (entry disc), the UKW (reflector), the battery
compartment and the light bulbs. It should be noted that the power knob and the
external power terminals are on the machine body
and protrude three cut-outs in the lid
when it's closed. This is different from later machines, such as
the Enigma I and Enigma K, were the power selector knob
and the terminals are fitted to the lid.
Compared to its predecessor — Enigma C — the following differences can be observed:
Immediately after the introduction of the Enigma D, several developments of improved machines were started. This led to a range of commercial and military machines, the first of which was introduced in 1927. The following machines are more or less based on the design of Enigma D.
 | ||||||
|
|
|
|
|
|
|
|
|
Machine with a single-ended Steckerbrett.
This eventually led to the developement of the Enigma I (Ch.11f),
which was used by the Reichswehr
(later: Wehrmacht) throughout WWII.
As far as we know there are no surviving examples of the Reichswehr Enigma D,
as the existing machines were all recalled and upgraded to Enigma I (Ch.11f).
From 1927 onwards, right up to 1944, this was the main commercial
(non-Stecker) machine. It replaced Enigma D and has several improvements.
This machine was also the base for the Enigma T (Tirpitz),
the Swiss Enigma K variant
and the Enigma KD.
This was the first of a range of Enigma machines that had a counter
and cog-wheel based rotor stepping mechanism. In addition, the rotors had multiple
turnover notches, resulting in an irregular stepping pattern.
The later Enigma G (G31, Ch.15a) is based
on the A28.
This was a numbers-only version of the Enigma machine. It had just 10 keys
(0-9), 10 lamps and the rotors each had 10 contact points at either side.
The simplified circuit diagram below, shows how the Enigma D operated.
At the right are the keys (switches), the lamps and the battery. The current
is supplied to the contacts of the right side of the first rotor, via the
entry disc (German: Eintrittswalze, or ETW). The current then passes
the three cipher rotors, each of which can be advanced by a key-press,
like the odometer of a car.
At the bottom right is the four-position rotary selector that acts as the
power switch. It allows selection between hell (bright), dkl (dark),
aus (off) and Sammler (accumulator). When set to dkl,
a series resistor (R) reduces the current through the lamps.
When Sammler (accumulator) is selected,
the voltage from an external battery or transformer can be applied
to the terminals (
At the far left is the reflector (German: Umkehrwalze, or UKW), which returns
the current through the rotors. The current leaves the rotors via the
ETW again, and causes a lamp to be lit. Note that the process is symmetric,
or reversable, meaning that the same settings can be used for coding and
decoding. This is a property of the reflector. Unfortunately it has the
unwanted side effect that a letter can never be enciphered into itself.
This is true for all glow lamp Enigma machines.
Enigma D has three cipher rotors and a settable reflector (UKW). All four rotors protrude the top lid of the machine. As the UKW has a thumbwheel for setting its position, it is nearly identical to a cipher rotor. As a result, the machine is sometimes erroneously classed as a 4-rotor Enigma, but is actually a 3-rotor Enigma with a settable UKW. The UKW does not move during encipherment. Each rotor has 26 positions and 26 contacts at either side, corresponding to the letters A-Z that are engraved in the outer ring. This index-ring can also be set to 26 positions (Ringstellung).
|
The rotors are made from steel and aluminium,
and have a Bakelite inner core with
26 contact pads at one side
and 26 spring-loaded pins at the other.
Each rotor can be set to 26 positions,
each of which is identified with a letter (A-Z).
As with earlier Enigma machines, the notch ring is attached to the body of the rotor (rather than to the letter ring). The image on the right shows the left side of an Enigma D rotor. The frontmost ring is called the notch ring. It has a triangular cut-out (notch), that is responsible for advancing the rotor to the left of this rotor. The notch ring has no mounting screws and is attached to the rotor body. As a result, the Ringstellung (ring setting) serves no cryptographic purpose, and does not contribute to the machine's key space. This can be seen as a design flaw. It is fixed in later machines like Enigma K and Enigma I, in which the notch is attached to the letter ring. |
|
This means that with Enigma K, the Ringstellung affects
the position of the notch relative to the wiring.
The image below shows the rotors sets of Enigma D and Enigma K
side by side. If we take a closer look at the rotors,
we'll see that
the rotors of Enigma K have horizontal screws
by which the notch ring is attached to the letter ring.
These screws are missing from the Enigma D rotors.
Distance between the rotors of Enigma D.
Although the diameters of Enigma K and Enigma D rotors are identical,
they are physically incompatible. The rotors of Enigma D are 2 mm wider
than those of Enigma K. Furthermore, the spacing between the rotors is
2 mm wider on the Enigma D. As a result, the total distance between UKW and
ETW is 12 mm higher on Enigma D, which means that the rotors are not
interchangeable.
Move the mouse over this image to see the rotors of Enigma K.
Below is the wiring of Enigma D and its rotors, as described in 1927
by Hugh Foss in his report
The Reciprocal Enigma [6].
The wiring is identical to the wiring of the later Enigma K (A27)
and can therefore be classed as 'commercial wiring'.
Note that on Enigma D rotors, the turnover notch is attached to
the rotor body, rather than the alphabet ring (index)
as on Enigma K rotors. This means that when the
ring setting is changed, the turnover of the adjacent rotor will happen at
a different letter. Furthermore, the position of the notch was identical
on all three rotors (I, II, III).
The table below shows the position of the notch when the ring is set to 'A'.
In that case, the notch is located at the position of the letter 'H', and
a turnover of the adjacent rotor will occur when the letter 'Z' in visible
in the window.
When the ring is set to 'B', the turnover will occur at 'A', etc.
Turnover = RingSetting - 1
| Rotor | ABCDEFGHIJKLMNOPQRSTUVWXYZ | Notch 1 | Turnover 1 | # |
| ETW | QWERTZUIOASDFGHJKPYXCVBNML | |||
| I | LPGSZMHAEOQKVXRFYBUTNICJDW | H | Z | 1 |
| II | SLVGBTFXJQOHEWIRZYAMKPCNDU | H | Z | 1 |
| III | CJGDPSHKTURAWZXFMYNQOBVLIE | H | Z | 1 |
| UKW 2 | IMETCGFRAYSQBZXWLHKDVUPOJN |
With the first Enigma machines, the operating instructions were no more than a few A4 pages created on a typewriter. With the arrival of Enigma D, and the later Enigma K, a professionally printed instruction manual was included, complete with photographs and a fold-out at the rear.
|
Original operating instructions are extremely rare, but it is known that
the contents of the booklet were revised several times, probably to reflect
the differences between Enigma D and K.
The image on the right shows an original A5-size instruction booklet that was probably supplied with an Enigma K. The first page of this booklet however, shows a photograph of an earlier model, which is most likely an Enigma D. Although it is a low-resolution photograph, it is clear that there are no numbers printed above the upper row of lamps on the lamp panel. |
|
The photographs in the fold-out at the rear of this booklet, show a different machine on which the numbers (0-9) are printed above the upper row of letters on the lamp panel. Likewise, the upper row of keys holds both letters and numbers. This is thought to be a property of Enigma K. Furthermore, the power selector is of a later construction, in which the knob was fitted to the lid. On Enigma D, the knob is fitted to the power selector and protrudes a cut-out in the top lid. ➤ Read the booklet
About the booklet
The front cover holds a stamp with the number 33. The meaning of this number is not known, but it could be the year in which it was issued (1933). In the upper right corner, the number A833 is written with a red pencil. This is probably the serial number of the machine it was issued with. As the machine with serial number A818 in the Crypto Museum Collection has been identified as an Enigma K, it is likely that A833 was also an Enigma K.
Enigma D (A26) is an extremely rare machine, as it was only in production during the course of 1926. Only a handful have survived, of which serial numbers A319 and A320 are described below. A320 is the machine that was acquired in 1926/27 by GC&CS (later: GCHQ) for analysis. We will deal with that machine further down this page, but we will first take a look at the A319.
The machine featured above – serial number A319 –
was found in an abandoned storage locker in the US. It was auctioned in
February 2024 and is now part of the Crypto Museum collection.
Little is known about the provenance of this machine, but from the other
items in the estate it seems likely that the previous owner was a
military officer who was stationed in Germany around 1958.
The machine was found with three rotors, numbered I, II and III, all with
matching serial numbers (A319).
All rotors have paper rings with handwritten numbers 1-26, glued to
both sides. This suggests that at some point the machine may have been used
for statistical analysis or even for cryptanalysis.
Although it seems likely that this machine was manufactured at the same time
as the A320 below, there are significant – cosmetic– differences, which are
further described below.
In 1926, Enigma manufacturer Chiffriermaschinen Aktiengesellschaft (ChiMaAG) had two machine offerings: (1) Die Schreibende Enigma (the printing Enigma) — a large, heavy and expensive non-reciprocal cipher machine — and (2) Glühlampenchiffriermaschine (glow lamp cipher machine) Enigma D — a portable light-weight reciprocal cipher machine that was developed as a cheaper alternative. As the former had been sold out completely and its successor was not expected for another year or so, it was suggested that the British Army might want to try the Enigma D [12].
|
British Naval Attaché in Berlin Lt. Hume took up the offer and
secured an Enigma D with serial number A320.
In late 1926 or early 1927, on a visit from his friend
Commander Edward Travis – deputy head of the
British codebreaking agency GC&CS (now: GCHQ) –
Hume passed the machine to Travis, who brought to the UK
for analysis.
It is rumoured that another Enigma D (A323) prompted the British Air Force to develop its own rotor-based cipher machine, which became known as RAF Enigma with Type-X attachments, commonly abbreviated Type-X or simply Typex. |
|
The machine was analysed in 1927 by GC&CS cryptanalyst
Hugh Foss, who reported his findings in
The Reciprocal Enigma
[6]. In his report, Foss describes the machine and suggest a method for
breaking it, assuming that a crib (known plaintext) of sufficient length
is available. 10 years later, in 1937, Foss' analysis was used by codebreaker
Dillwyn (Dilly) Knox, to develop two methods for breaking Enigma: Rodding
and Buttoning Up. It allowed codebreakers William Bodworth and Dilly Knox
to break the Enigma K of the Spanish Navy and the Italian Navy respectively,
in 1937 [12].
On 14 September 2021, GCHQ issued a Tweet about the visit of veteran
codebraker Charlotte 'Betty' Webb to their offices at Cheltenham [7].
One of the photographs (above)
shows Betty Webb in front of a display with
two Enigma machines. Although these are a commercial Enigma K and a
Japanese Enigma T (Tirpitz), a barely readable text card in between
the two machines, provides some details about the A320 and
Hugh Foss' analysis of it: 1
Despite the fact that the above two machines have consecutive serial numbers (A319 and A320) and were probably manufactured around the same time, there are several cosmetic differences. So far, the following differences have been observed:
On the A319 the leather carrying grip is
at the rear,
whilst on the A320 it is at the right side panel.
Looking carefully at the
wooden enclosure, the case has
pre-drilled holes at both locations.
On the A319 the
holes at the side have been stuffed.
The metal cover of the Enigma D, has
four oval holes
through which the current position of the three rotors and the UKW can be
observed. On the A319, these holes are covered by
shiny panes with a circular cut-out at the centre.
The cut-out is covered with transparent celluloid.
These window panes and their mounting screws, are
missing from the A320.
It is possible that they were originally present but were removed for some
reason.
On the A319, white numbers are visible on the
top row of the lamp panel,
and also below the keys.
These numbers are missing from the A320.
It is believed that the numbers were introduced on Enigma K (A27),
and were originally not present on Enigma D.
After close examination of the A319, it has become clear
that these numbers are handpainted (rather than screen printed).
From the discolouration of the white paint,
it is obvious that the numbers were added many years ago,
probably when the machine was issued in 1926.
Apart from the differences, there are also similarities:
A more likely explanation is perhaps that the knob was removed deliberately by
the original user, as it was a nuisance to place it in the
north/south position – in
between two settings – each time the lid had to be opened. This is especially
true when the machine is used for analysis, which was the case with the
A320 and probably also with the A319.
The black knob of the power selector is missing from both machines
(A319 and A320).
It is narrower than that of the later Enigma K, and
protudes a cut-out in the metal lid.
As it was probably cast in a light metal alloy, and has a 4 mm hole to
fit the axis of the power selector,
it might have been fragile, as a result of which it could have been broken.
The serial number is
stencilled to the front of the wooden case.
The serial number is also impressed into the bottom of the machine's
base plate (near the rear edge),
on the lamp panel mask (inside the machine), and
at the right side of the cipher rotors and the UKW.
The Enigma D is held in the wooden case by four large bolts at the bottom of the case. After removing these four bolts, the machine can be lifted from the case. Unlike the later Enigma K and Enigma I, Enigma D does not have a metal case shell. The wooden case acts as an enclosure.
|
This is particularly visible when looking inside the battery compartment
before lifting the machine from the wooden case, as shown in the image
on the right. There is no box to hold the battery, just four metal brackets
at the edges.
Also visible in the image on the right are the power selector and the external power terminals which are permanently mounted to the machine body. They protrude the top lid through three cut-outs. This is different from later machines, were they are mounted to the lid. Note that the knob of the power selector is missing here. |
|
The image at the top of this section shows the bare machine after removing it from the wooden case. At the bottom is a large steel plate that acts as the machine's base. Just above the base plate is a horizontal bar – made of bended steel – that actuates the wheel stepping mechanism when a key is pressed. In later models, the base plate and the bar were cast in a light metal alloy.
|
Enigma D is the first machine with a hinged top lid, which makes it
easier to access the rotors and alter the settings. The lid is held in
place by two retaining bolts, just above the keyboard.
The image on the right also shows the construction of the lamp panel. It consists of a metal plate with 26 circular holes – one for each letter of the alphabet – to which 26 rings are soldered. It acts as a mask for the 26 light bulbs of the lamp panel and the alphabet film (A-Z). In later machines, this was also a cast in light metal. Also visible in the image are the rotor windows. |
|
Although the cut-outs in the top lid are oval shaped, the actual window panes – mounted at the upper side of the lid – are circular. Note that these circular window panes are not present on the A320. It is possible that they were originally present, but were removed for some reason.
|
Opening the lid provides access to the machine's interior, which
consists of the lamp panel, the battery compartment, the power selector
and the rotors. The rotors can be released by putting the
rotor locking lever in the
upright position, after which the UKW can be
shifted to the left and the three cipher rotors –
fitted to a common spindle
– can be lifted from the machine. This reveals the stepping mechanism
as shown in this image.
The UKW (on the left) can be removed from its axle. The ETW (on the right) however, is permanently wired and can therefore not be removed. |
|
The rotor stepping mechanism consists of
three spring-loaded stepping levers,
that are actuated by the large stepping bar underneath the keyboard.
Each stepping lever has a triangular-shaped tip at the top, but the rightmost one
is narrower than the other two. It drives the sawtooth ring (ratchet) of the
rightmost rotor, as a result of which it advances by
one step on each key press.
The other two stepping levers move in tandem
with the rightmost one, but can only drive the corresponding rotor
when its triangular tip engages with the
triangular cut-out (i.e. the notch)
of the rotor to its right. This principle is known as Enigma Stepping,
and has the inherent property that the middle rotor can make an extra
step on two successive key presses.
This so-called
double stepping anomaly
was described in 1997 by David Hamer [13].
It marginally reduces the machine's
key space and the cipher period, but the effect is not as dramatic as it may seem.
June/July 2024
The Enigma D with serial number A319 – featured on this page – was acquired
at an auction in Florida (USA) in February 2024. Judging by the photographs
of the auction house, it was in good condition – especially for its age –
but it was uncertain whether it would be in working order.
| As soon as it arrived, we closely inspected it. The case was complete, but had green stains at the front, probably from a leaking paint can. It was clear that this had happened many years ago, probably when the machine was stored in an attic or garage. Furthermore the lock was badly damaged and the mating pin – which normally resides in the case lid – was missing. We later learned that it was broken off and got stuck in the lower part of the lock, where it blocked the lock's mechanism. Apparently someone had attempted to open the lock with excessive force, as a result of which it was now badly damaged. |
|
The machine itself was in very good condition, especially if you realise that in two years time it will be 100 years old. We already knew that the power knob was missing and that the lamp film had deteriorated and shrunk, to the point where some of the letters were barely visible.
|
It was decided to replace the lamp film with an
accurate replica that
we created especially for this machine. The original film will be
kept as proof of its genuine origin,
but is safely packed. Acetate films
of this type may emanate potentially harmful gasses, even after so
many years.
The machine itself appeared to be in excellent condition. After a bit of superficial cleaning, replacing a couple of broken and missing lamps, and inspecting the battery contacts, we used the power supply unit (PSU) in our lab to apply 4.5V DC with a constant current of 200 mA. |
|
Most keys and lamps worked straight away, even without cleaning the rotor contacts. Again, not bad for a machine of this age. All three cipher rotors had paper rings with numbers (1-26) glued to both sides, as an index to the contacts. We assume this was done for some kind of analysis. Judging from the paper, the ink, the handwriting and the glue, this appears to have been done a long time ago, probably shortly after the machine was purchased in 1926. As these paper rings are part of the machine's history, we decided to leave them in place and treat them carefully.
|
It was decided not to treat the green paint stains at the front of the
wooden box, as in that case we would also loose the original serial number which
is stencilled at the front, just below the lock.
The lock was removed and,
with help from Austrian collector Günter Hütter, taken apart, reworked,
straightened and reassembled.
Even a new mating pin
for the case lid was machined.
Whilst the machine was outside the wooden case, its side panels were rejoined with a bit of fresh glue. At the same time, the broken leather carrying strap was replaced by a good reproduction. |
|
|
Scratches near the cut-out for the knob, reveal that previous owners
made a mistake with this more than once.
In later models, like Enigma K and Enigma I,
this was solved by
permanently fitting the knob to the top lid,
and driving the power selector indirectly with a
notched disc.
Another weakness in the design of the power selector is the metal plate shown in the image on the right. It covers the shafts of the power selector and the two external power terminals. It is held in place by springs around the shaft of the power selector and around the power terminals. They are used to press the plate against the lid. |
|
Unfortunately, the plate and the springs are made of metal and can therefore cause a short-circuit between the shaft and the power terminals. To prevent his, a piece of Pertinax (Paxolin) is fitted to the bottom of the plate. In our case however, the springs had been deformed somewhat, as a result of which they could bypass the pertinax and make contact with the metal plate above it. This was solved by reshaping the springs of the power terminals and replacing the other one.
Device Brand Type Model Designator Manufacturer Country Year Customers Predecessor Successor Descendants Rotors Contacts Turnovers Reflector Wiring Stepping Plugboard Dimensions Weight Quantity
Enigma D is known under the following names:
A316 A317 A318 A319 A320 A323 A324 A325 A326 A327 A328 A338 A343 A344
A319 A320 A324 A327 A343
The following people have contributed to this page and/or the restoration of the A319:
ChiMaAG, 1933. Issued with Enigma A833.
Enigma serial number research. Forthcoming publication.
September 2022, February 2024.
Instruction booklet with images.
September 2022.
Bletchley Park, Security Conference, 2010.
ISBN 978-3540668718. 1 January 2000 (2nd edition). p. 113.
TNA, HW25/14. Undated, but probably 1927/28.
Brief account of the purchase of A320 and
Hugh Foss' analysis.
Twitter, 14 September 2021.
Transcription of barely readable A320 description in GCHQ display case [7].
15 September 2021.
Swedish Views on Enigma Development 1924-1930
Histocrypt Proceedings, 15 May 2018.
In reference to a letter from AB Cryptograph,
dated 5 April 1927 (Swedish).
Paris (France), 11 April 1927.
Preliminärt utlåtande angående "Glühlampen-Chiffriermaschine Enigma" (Swedish).
Paris (France), 1 August 1927.
1949. Published in Chapter 3 of the book Action This Day.
Michael Smith & Ralph Erskine, 2001. ISBN 978-0-593-06357-6.
David H. Hamer.
Cryptologia, Volume XX, Number 1, January 1997. Author's copy.