The authors assert that diagnosis based upon ordinary psychiatric interviewing is notoriously inaccurate, and so they set out to devise certain tests and to adapt tests which have been used somewhat differently in order to have at hand a relatively infallible method of diagnosis of psychiatric illness and in particular to differentiate between normal people, neurotics, and psychotics. The results turn out to be even more favourable than the authors expected. By such methods they hope to eliminate the necessity for unscientific and intuitive approaches upon which they pour disdain, albeit relatively gently. The only statement which, if taken seriously, would put every psycho-analyst up in arms is in the last paragraph of the book. To quote: 'Most existing theories may be adequate to deal with highly abstract concepts having little or no anchorage in behaviour, but they do not enable us to make predictions in the field covered