A list of misprints will appear here shortly.

For the moment I just mention the most important correction.
The proof of the crucial Lemma 10.89 starts with an application of Lemma 10.88. However, the formulation of Lemma 10.88 is too weak to give the desired conclusion. I am very grateful to Ian Hodkinson for pointing this out.

Lemma 10.88 should be stated in a stronger form in order for the the proof of Lemma 10.89 to work.