Why did the Germans use the Enigma machine rather than the far superior "Lorenz" cipher machine?

Another point: You are judging the situation from the modern perspective, in which people have learned what makes an encryption method secure and what makes one insecure. People at the time of Enigma etc were much less knowledgeable about this. For them, 'security through obscurity' was naively considered a valid way of thinking about encryption, whereas now we understand that it is really a dubious way of thinking. Actually, it was experience in breaking codes such as Enigma that helped people learn these things.

[CLARIFICATION: In the preceding paragraph, I wasn't actually claiming that -Enigma itself- used 'security through obscurity' (though the ways it was set up for use showed this feature to a limited extent); I was merely noting that -that general approach- was a common mindset (and often still is, to our detriment). I apologize for the confusion.]

Additionally, with regards specifically to Enigma: Aside from the built-in flaws of the scheme itself, breaking the code was greatly aided by some pretty bad practices in the way the Germans used the Enigma, which failed to concern them precisely because they assumed it was unbreakable. [for example, one Enigma operator chose his encoding gears by the fact that their letter identifiers matched the initials of his favorite Hollywood movie stars (or something like that); this helped the British decode his messages once they picked up on this fact]

"Superior" is a subjective attribute. To use that term, you basically have to define it (faster, more reliable, etc.)

What appears to be true about the Lorenz machine is that it is more sophisticated than other machines such as Enigma. Another way of putting this is that it was more "systematic," and therefore more efficient in its workings that the Enigma machine. That's one way of defining "superiority."

That can also be a disadvantage. If a code is "too" systematic, it can be easily broken. This appeared to be the case with the Lorenz machine; British crytographers were able to reconstruct its structure before the war ended without having seen it in action, and post-war investigators found it easy to break. Enigma, by its complexity, was actually more secure (and easier to make secure).

In one method of coding messages, America used Navajo Indians (and their relatively arcane language) on both sides of the messaging system. These are human codes, but one that is non-systematic and therefore relatively unintelligible, unless you have a Navajo to decipher it for you.

I would add two more factors: cost and being invented too late.

Enigma was available commercially in 1923. The Reichsmarine (the navy of the German Republic) put it into service in 1926 and the Reichswehr in 1928 (the army of the German Republic). This meant by 1939 the German military had 10 to 15 years experience with Enigma, and German industry had experience producing it.

The Vernam cipher, upon which Lorenz is based, was invented in 1917 and some machines existed in the 1920s. But the Lorenz SZ40 didn't exist until 1940 and was not brought into operational use until 1941.

Enigma machines were relatively cheap, portable, robust, and required no external power. It's estimated 20,000 to 50,000 were produced (I don't know if this is military machines or all machines). This allowed them to be used tactically by individual ships, submarines, and division commanders. 1,100 were required for submarines alone.

Switching the German military to Lorenz for tactical communications could only happen in 1941 the earliest. It would have required producing thousands of the new, complex machines at a time when Germany was being drained of resources by garrisoning their newly conquered empire and their invasion of the Soviet Union. New protocols in using the machine securely at the tactical level (operational mistakes are what sunk Enigma and would also sink Lorenz) would have to be developed and distributed. Thousands of operators would have to be retrained on how to use the machine.

Distribution and training would have to happen from Atlantic Ocean to the depths of Russia, from the deserts of North Africa to the fjords of Norway. Unlike Enigma which was worked out in peacetime, the German military would have to learn this new device while in the middle of a high-intensity, multi-front war.

All to be cracked in about a year or two.

As far as I can tell, they never managed to break the Lorenz machine.

Tommy Flowers would be very surprised to hear that. He lead the development of machines to crack Lorenz culminating in Colossus, arguably the first programmable digital computer.

Basically, the only good thing about the "Enigma" cipher machine was its name. Other than that, it was grossly inferior to the "Lorenz" one, apparently only used by some of the main, top-ranking persons.

If anything Lorenz prove easier to crack while Enigma continued to give code breakers headaches throughout the war.

Enigma was being analyzed since the 1920s, first by the Poles, then by the British. Physical copies of the machines were available to the code breakers. The Enigma machines were continually improved with features such as the plugboard and additional rotors sending cryptographers scrambling to adapt.

In contrast Lorenz was first seen in late 1940. It was worked out without ever seeing a machine. Breaks appeared in January 1942. By mid 1942 Lorenz was being broken on a regular basis. 1943 saw machines and computers dedicated to breaking Lorenz.

Sources

• Crypto Museum
• Enigma FAQ by Dr Mark Baldwin
• Cryptanalysis of the Lorenz cipher

The Enigma was portable. You could carry one on a small submarine, or in an armoured battalion headquarters, and they weren't a problem when an airfield had to be moved in a hurry. They didn't require mains electricity, or special communications hardware - messages were sent and received by hand using Morse code - and they were believed to be secure, provided they were used properly, and the key lists and rotors were kept out of the hands of the enemy. It was also economical to procure, and the basic technology was well-proven.

By contrast, the Lorenz did require mains electricity and special communications equipment, and was only really suited to a permanent headquarters. It was known to be more secure, but the Enigma was thought to be sufficiently secure for everything except the highest level communications, so that's what the Lorenz was used for by the German Army. It was also expensive, and complicated by the standards of the time, which made it only usable in situations like a permanent headquarters, where skilled technicians could be available.

The Lorenz was also decrypted by the Allies, from mid-1942, after considerable amounts of previous work. It is not as widely known as the Enigma decryption, because the Enigma has a cooler name, and was much more tactically important. The Lorenz gave strategic-level information, which appeals less to journalists and movie-makers.

The Siemens and Halske T52 family were another family of enciphered teleprinters, used by the Luftwaffe and Kriegsmarine. They weren't compatible with the Lorenz. The a, b and c versions were weak enough to be broken by hand, but the d and e versions were stronger than the Lorenz. The British cryptanalysts at Bletchley Park broke into these ciphers, but didn't put as much effort into them as the Lorenz. This was because the Luftwaffe often re-transmitted messages in ciphers that were already broken, or much easier to attack. Also, reading the Lorenz used by the German Army would usually tell you what the Germans intended doing.