How did the Bletchley Park code breakers figure out the chi stream of the Lorenz cipher, that was obscured in the ciphertext, which British code breakers eventually decoded? It's written in The Rutherford Journal that John Tiltman deduced the key using a depth found on 30 August 1941. Since he couldn't figure how the key was generated, it was given to W.T. Tutte, who figured out how the key was generated without ever looking at the Lorenz cipher machine.
Jack Copeland writes in Colossus: The Secrets of Bletchley Park's Codebreaking Computers (p.364)
Here was a method of wheel setting! Δ(χ1+χ2) was a periodic sequence, supposed known, with a period of 41×31=1271. Lay it against Δ(C1+C2) in each of the 1271 possible relative positions and count the number of agreements for each one.
It's obvious that the British code breakers had the ciphertexts in hand, from which they could obtain the delta-ed version of the sum of the first and second impulse of an intercepted ciphertext, i.e., Δ(C1+C2). I would like to know how the code breakers figured out the delta-ed version of the chi stream of the Lorenz cipher machine, i.e., Δ(χ1+χ2).