by DAVID KAHN
The Abwehr did not have its own codebreakers. For such intelligence it relied, as part of the Wehrmacht, upon the military cryptanalytic agencies.
Of these there were four: one in the Oberkommando der Wehrmacht for the armed forces as a whole, and one each for the high commands of the Army (O.K.H., or Oberkommando des Heeres), the Navy (O.K.M., or Oberkommando der Kriegsmarine), and the Air Force (O.K.L., or Oberkommando der Luftwaffe). They traced back to an intercept and cryptanalytic service established in the Army in 1919 by one Lieutenant Colonel Buschenhagen, who had worked in the intercept service in the war. He called it the “Volunteer Evaluation Office” and installed it on the Friedrichstrasse. In February of 1920, its twelve-man staff moved to the Defense Ministry Building on the Bendlerstrasse, where it became Group II of the Abwehr. Since the unit’s work was much more closely allied with communications, however, it reverted a few years later to the administrative control of the chief of signal troops.
Even before that, it had moved out to nearby Grunewald, disguising itself as a newspaper translation and study group to avoid the Inter-Allied Military Control Commissions, which had proscribed intercept and codebreaking activities for the postwar German Army and had very nearly discovered the unit’s real activity. The unit further evaded both the letter of the Commissions’ directives and the spirit of the disarmament clauses of the Versailles Treaty when it began to prepare itself for any sudden demand for cryptanalysts that might arise—as in the case of a war. On October 31, 1921, the Army high command sent out a secret circular:
In order to cultivate and develop further the study of cryptography and the utilization of the results of the Intercept Service [Horchdienstes], it is necessary to train suitable officers for this special service branch.
Such officers are required to have a good knowledge of radio technique and mathematics, as well as geography, and some knowledge of a language (English, French, or an Eastern language).
The officers concerned are not to be detached. At first it is intended that instruction will be by correspondence only, using problems given by the Army command for the winter half-year.
Officers who distinguish themselves by especially good performance will be considered for service in evaluation stations of the higher commands and of the Army command; in addition there are expected to be prizes in the form of books on special branches of science.
The communications-intelligence unit stepped up its activities as Allied supervision waned. Part of its work consisted of picking up press association messages and news broadcasts and distributing a digest of them to government officials. By 1926, it had intercept stations in six major cities of Germany. In 1928, it began following the military maneuvers in which neighboring countries were once again engaging. It sneaked its intercept units into the demilitarized zone along the Rhine by disguising them as technicians for the German broadcasting or postal organizations. Much of its success resulted from traffic analysis—in 35 of the 52 major maneuvers between 1931 and 1937, the foreign forces were reconstructed completely. But it also solved some cipher systems.
When in 1934, Hitler pointed Germany toward its eventual war of revenge and conquest, he swelled the ranks of the armed forces and intensified military activities. But though the cryptologic agencies likewise grew in size, they did not necessarily grow in effectiveness. There were too few specialists in this recondite field to fill the need created by the proliferating military and party organizations. Some of the Army cryptanalysts were siphoned off to serve in the Forschungsamt, others, the Luftwaffe. Some of the intercept people moved over to Josef Goebbels’ Ministry of Propaganda, where their news-eavesdropping could help. About 1937, the O.K.W. created its own communications and cryptologic staff, thereby draining off more of the experts and further splintering the effort in the field. These new agencies were staffed by World War I veterans who were now rejoining the German Army; most had been officers in the signal corps but had no great experience in or aptitude for intercept or cryptologic work. By mid-1939, the German communications-intelligence services had 18 times as many people in them as they had had in 1932, but useful results had in no way kept pace.
Six days before Hitler fell upon Poland, Major General Erich Fellgiebel, 52, who had been in communications since he joined a telegraph battalion upon enlisting in 1905, was named head of the O.K.W. communications organization. His title was Chef, Wehrmachtnachrichtenverbindungen (“Chief, Armed Forces Signal Communications”), or Chef W.N.V. His superior was the O.K.W. chief, Field Marshal Wilhelm Keitel, whose only superior was Hitler. Keitel wrote in Fellgiebel’s fitness reports: “In his field a pronounced leader type with foresight, a gift for organization, full energy and dedication…. In his attitude towards National Socialism an inclination to unconsidered overcriticism….” The W.N.V. supervised communications, including communications security, and intercept operations;* it served as a kind of staff, an advisor and controller, for the service branches that largely operated the communications and intercept networks for the Army, Navy, and Air Force, much as the O.K.W. itself advised and directed the service commands.
Under the Chef W.N.V. came the Amtsgruppe W.N.V. Its chief was Major General Fritz Thiele, 48, a close colleague of Fellgiebel’s who had previously headed the O.K.H. communications and intercept organization. He became Chef, Amtsgruppe W.N.V. the day the war began. The unit comprised radio and wire branches, which maintained communications between the headquarters of the three armed forces high commands, a technical equipment office, an administrative office, and the Chiffrierabteilung (“Cipher Office”), usually abbreviated “Chi.” Colonel Siegfried Kempf assumed command of Chi on the same day that Fellgiebel became Chef W.N.V. Then 43, he was a career communications officer, a martinet disliked by his subordinates. He was succeeded in October, 1943, by Colonel Hugo Kettler, 48, who had had considerable intercept experience and who brought out the best in his men.
In 1944, the Chiffrierabteilung was divided into eight groups. Four came directly under Kettler; the other four were combined into two supergroups, Gruppen II and III into Hauptgruppe A for cryptography, Gruppen IV and V into Hauptgruppe B for cryptanalysis, each with its own head who reported to Kettler. This was the organization:
Gruppe Z (Zentralgruppe): personnel; pay, administration; office space and furnishings; Nazi ideological supervision.
Gruppe I: Organization and Control. Referat Ia: direction of the international monitoring service (Chi had intercept posts in Madrid and Seville as well as Lörrach and Tennenlohe, with main posts in Lauf and Treuenbrietzen). Referat Ib: study of foreign communications systems. Referat Ic: provision of teletype communications for Chi and R.S.H.A./VI/Mil (former Abwehr).
Gruppe II: Development of German Cipher Methods and Control of Their Use. Referat IIa: camouflage methods for telegraph and radio messages; intercept and wiretapping techniques; cryptographic policy; supervision of cipher employment; cryptographic compromises. Referat IIb: development of German cipher systems (camouflage methods, secret writing, secret telephony); supervision of and instruction in cipher production. Referat IIc: cryptographic systems for radio agents.
Gruppe III: Cipher Supply. Control of production, printing, and distribution of ciphers and keys; operation of the distribution posts (headquarters at Dresden with depots in Halle, Zwickau, Chemnitz, Leipzig, Frankfurt-am-Oder, Bischofswerda, Magdeburg, and Reichenbach).
Gruppe IV: Analytical Cryptanalysis. Referat IVa: testing of suggested German military cryptosystems and telephone scramblers for resistance to cryptanalysis: examination of inventions. Referat IVb: development and construction of cryptanalytic apparatus for Wehrmacht cryptanalytic units; operation of the equipment at Chi. Referat IVc: development of cryptanalytic methods; stripping of superencipherments for Gruppe V. Referat IVd: instruction.
Gruppe V: Practical Cryptanalysis of the Messages of Foreign Governments, Military Attachés, and Secret Agents. Referate V 1-22: national offices. Referat Va: Wehrmacht codewords.
G
ruppe VI: Interception of Broadcast and Press Messages. Referat VIa: radio reception technique; administration and control of the listening posts at Ludwigsfelde, Husum, Münister, and Gleiwitz. Referat VIb: interception of radioed press and teletype transmissions and of international radio traffic. Referat VIc: surveillance of transmissions from within Germany to the outside. Referat VId: evaluation of broadcasts and press communication; issuance of the Chi-Nachrichten (a 10- to 20-page daily summary of the noncryptographic intercepts); special reports.
Gruppe VII: Referat VIIa: evaluation and distribution of output. Referat VIIb: chronicles of events (perhaps serving as an information unit).
In addition to these eight sections, a working committee for the testing of German cryptographic security reported directly to Kettler, and half a dozen intercept companies worked for Chi. The office was expected to maintain liaison with the communications units of the Army, Navy, and Air Force; with the chief of army equipment and commander of the replacement army, under whom there was an inspector of signal troops; with the R.S.H.A., the Foreign Office, the Propaganda, Post, Air, Trade, and War Production ministries and, of course, with the party.
(By 1945, the Chiffrierabteilung had been reorganized into seven groups, with functions apparently as follows: Gruppe Z, administration; Gruppe I, organization and control; Gruppe II, Chi-Nachrichten; Gruppe III, broadcast and press interception; Gruppe IV, cryptanalysis; Gruppe V, teletype for Chi and R.S.H.A./VI/Mil; Gruppe X, evaluation, distribution and information services. This downgrading of cryptanalysis and upgrading of the non-cryptanalytic results may reflect a drop in the cryptanalytic results late in the war.)
Chief of Hauptgruppe B, in which the cryptanalytic functions reposed, was Ministerial Counselor Wilhelm Fenner, 48 when the war started. A German born and raised in St. Petersburg, he had headed German military cryptanalysis since 1922. He was a brilliant organizer who oversaw the expansion of the group from a handful to more than 150, but he handicapped himself by his egocentricity and by his superciliousness with regard to the noncryptanalytic aspects of communications intelligence. His right-hand man was a Russian emigrant, Professor Novopaschenny, who under the Czar had been attached to an astronomical observatory in Pulkovo, outside St. Peterbsurg. He developed much of the technical aspects of the work, but seems to have held only a relatively subordinate post as a chief cryptanalyst in one of the national offices, apparently Referat V 9, which was probably Russia.
Head of Analytical Cryptanalysis (Gruppe IV) was Dr. Erich Hüttenhain, who also directed that group’s instructional activity (Referat IVd). Referat IVa, which tested German cryptosysterns, frequently with mathematical tools to calculate theoretical limits of security and to find improvements, was headed by mathematician Dr. Karl Stein, who held the rank of lieutenant, surprisingly low for so lofty a position. Referat IVb, headed by Engineer Wilhelm Rotscheidt, used tabulating machines and special-purpose devices. It invented the prototype of the translucent-sheet-and-light device used by Pers Z to strip additives from a known code. The unit first worked out the device for two-digit codes and then extended it to four. Instead of punching out holes corresponding to the most frequent groups, however, Referat IVb marked them with small crosshatched disks, and looked, not for the brightest spot, but for the darkest. Stein’s mathematicians extensively investigated the question of how a codegroup stock could be constructed so that this method would not work against it. Referat IVc’s chief was Professor Dr. Wolfgang Franz, and Ministerial Counselor Dr. Victor Wendland was head of Gruppe V (Practical Cryptanalysis) and so Fenner’s immediate subordinate.
Early in the war, the O.K.W. cryptanalysts worked in a former town house on one of the streets that run off the Tirpitzufer, not far from O.K.W. headquarters on the Bendlerstrasse. About 1943 they moved to much larger quarters in a modern semicircular concrete office building at 56 Potsdamerstrasse called the Haus des Fremdenverkehr—a name that gave rise to many bad jokes because “fremdenverkehr” (“tourist traffic”) is German slang for “fornication.”
On July 21, 1944, Fellgiebel’s sudden removal from command rocked the whole W.N.V. It seemed to be connected with the bomb attempt on Hitler’s life of the day before—and it was. Fellgiebel, whose anti-Nazi proclivities had been noted in his fitness report by Keitel, had in fact been a key figure in the plot. He was replaced by Thiele, who became head of both the O.K.W. and the O.K.H. agencies. He served for exactly a month. Then he was arrested as a co-conspirator, his personnel file crossed out with a giant X, and the entry made under his name, “stricken from the honor roll of the German Army and the Wehrmacht!” Fellgiebel had been executed on August 10; Thiele soon followed. Lieutenant General Albert Praun took Thiele’s place in both offices and retained them to the end of the war.
The oldest, most experienced, and closest to O.K.W. of the other cryptanalytic agencies was the Army’s Heeresnachrichtenwesens (“Army Communications System”), or H.N.W. The Chef, H.N.W., served on the Army general staff. Like the U.S. Army’s Signal Corps during World War II, it had both communications and intercept-cryptanalysis duties; like the Signal Corps, it turned over its solutions to Army intelligence for evaluation and use.
Under Chi’s watchful eye, it issued cryptosystems for the troops. For high-level communications, from the O.K.H. down to regiments, the Army used the glowlamp Enigma cipher machine. It was reliable, working well in the Russian winter and the Libyan summer. Signal officers thought it cryptanalytically secure if—as ordered by 1942—keys were changed three times a day. Its chief disadvantage was that it did not print its output. Battery-powered and portable, it could be operated in a moving truck and was well adapted to radio work.
Nevertheless, in 1943 a new machine began replacing it in some areas. This was a printing machine, produced by the Wanderer Werke firm, which copied the Hagelin variable-gear principle. There is a story that one of these was found in Norway at the end of the war with a message still in it, obviously abandoned by an operator who disagreed with what he had deciphered: Der Fuehrer ist tot. Der Kampf geht weiter. Doenitz (“The Führer is dead. The war goes on. Dönitz”).
For wire teletypewriter communications from the O.K.H. to army corps and a few divisions, the Germans used an on-line machine produced by Siemens & Halske Aktiengesellschaft. Its heart was a set of ten keywheels, similar to those on a Hagelin machine, rimmed with pins that could be made either operative or inoperative. Each wheel had a prime number of pins, ranging from 47 on the smallest to 89 on the largest. Five of these wheels enciphered the five teletypewriter pulses, transforming a mark into a space or vice versa if the pin then in position was operative, or leaving the pulse unchanged if it was inoperative. The other five wheels effected a transposition of the pulses. The machine enciphered and transmitted in a single operation, and likewise deciphered and printed out the message automatically.
Beginning in June, 1942, regiments, battalions, and companies enciphered with the double transposition, with the same keyword for both blocks—the same system, interestingly, as the German Army used at the start of World War I. (This system also backed up the Enigma.) Each division produced at least three keys for its subordinate units. The troops heartily disliked the double transposition, however, and cleartext messages showed a noticeable upsurge. For intelligence and combat reports, these units used small three-letter or three-digit codes, which were likewise published by their divisions. Many cipherers preferred their simplicity to the complexity of the double transposition, and often used them for orders and other unauthorized messages. A signal officer complained bitterly of this practice: “Tarntafeln sind kein Schlüsselersatz!” (“Code tables are not cipher substitutes!”), he wrote in a report. Later in the war, a bigraphic substitution replaced the double transposition as a front-line cryptosystem, and in 1944 a modification of the grille replaced that. In addition, the signal troops used numerous special ciphers—for call-signs, numbers, and so on.
The H.N.W. communications-intelligence service operated as a separate or
ganization within an army or an army group, though parts of it were sometimes specially assigned. In 1943, for example, the commander of Fernmeldeaufklärung 7 (“Radio Intelligence 7”), reported to Field Marshal Albert Kesselring. Fernmeldeaufklärung 7 consisted of radio intelligence companies and platoons and direction-finding stations widely scattered over the central Mediterranean area—in western Crete, southern France, northern Africa, Sicily, Sardinia, and Italy. These units reported their intelligence results via their own radio net to the headquarters at Rocca di Papa, south of Rome; the original intercepts were then forwarded to headquarters for more comprehensive evaluation. Fernmeldeaufklärung 7 distributed radio intelligence of tactical importance to the lower commands by broadcasting it in a special cipher. While much of this intelligence came from conversations or radio messages in plaintext or from traffic analysis, much also came from cryptanalysis. Similar units on other fronts also provided valuable material.
Thus when Hitler, in a fit of rage, fell upon Yugoslavia, his armed forces overran that tough little nation across mountainous terrain formerly considered blitzproof with a speed that could not be fully accounted for even by their overwhelming strength. And in fact the Germans could exploit the Yugoslav military messages to tell their Panzer commanders where and how their armored columns might best spear down toward Zagreb, Sarajevo, and Belgrade. For since January, 1940, German Army intercept personnel, wearing civilian clothes, had monitored Yugoslav emissions from an intercept station in Sofia and had broken the Yugoslav military cryptosystem.
After the conquest, a radio intelligence platoon cryptanalyzed the ciphers of the partisans under General Draja Mikhailovich and his Communist rival, Tito. The results enabled the occupation troops to forestall many guerrilla depredations. Tito, finding some of his efforts frustrated, at first suspected treachery and purged some of his underlings. Soon, however, he guessed the truth and changed his ciphers with great frequency but with no success. In the spring of 1943, for example, the Germans picked up a series of messages from which it became clear that Tito’s relations with his Anglo-American allies had deteriorated. Others spoke of a proposed Anglo-American landing on the Adriatic coast—but this never came off.
