When I take up a new book I always read its conclusions first. For I have discovered that even difficult books have conclusions that are easy to read and that give a first idea of their content. Of course, I expect my readers to do this also! Indeed, the present book offers them an opportunity for generalizing the method. After having gone over chapter 14 they should preferably look through chapters 5, 2, 13 and 10, or at least some of these (in this or in other orders), before engaging into any thorough reading. The point is that-at appropriate places in other chapters-the reader will find proofs; and, as we all know, all the proofs that are worth something are unavoidably ponderous. Of course, it is good to have them but, again, the most suitable procedure not to lose interest in their development is, I think, to first provisionally grant their conclusions,. in order to see where they lead us to. Then, to study the proofs once we have been convinced that the goal is worth some effort. This is the reason why I suggest that at least chapters 4 and 11 should, on a first reading, be skipped, at any rate, by the "poets"! The problem, a new approach of which is described in chapter 4, did not come to light, at least in its present form, before 1964. The first experi­ mental investigations bearing on it date back from the seventies.

When I take up a new book I always read its conclusions first. For I have discovered that even difficult books have conclusions that are easy to read and that give a first idea of their content. Of course, I expect my readers to do this also! Indeed, the present book offers them an opportunity for generalizing the method. After having gone over chapter 14 they should preferably look through chapters 5, 2, 13 and 10, or at least some of these (in this or in other orders), before engaging into any thorough reading. The point is that-at appropriate places in other chapters-the reader will find proofs; and, as we all know, all the proofs that are worth something are unavoidably ponderous. Of course, it is good to have them but, again, the most suitable procedure not to lose interest in their development is, I think, to first provisionally grant their conclusions,. in order to see where they lead us to. Then, to study the proofs once we have been convinced that the goal is worth some effort. This is the reason why I suggest that at least chapters 4 and 11 should, on a first reading, be skipped, at any rate, by the "poets"! The problem, a new approach of which is described in chapter 4, did not come to light, at least in its present form, before 1964. The first experi­ mental investigations bearing on it date back from the seventies.