appeared during his attacks, we are safe in assuming that one factor at least in the production of an attack was the lifting of some inhibition which kept the cruder ideas from entering consciousness except in a form in which they could be objectively viewed and so logically arranged. Scattering of thought therefore arises from the intermittent action of this censor or from an incomplete abolition of the inhibition allowing varying formulations of the crude ideas to gain expression which have no logic surface connection. If entirely done away with, of course, the latent ideas appearing in perfect crudity would have a logical connection. The content of consciousness is what is within the sphere of introspection. We can therefore say that the praecox who is scattered really does not know his own ideas. This is, of course, an every day experience for those who examine such patients. A suitable case left to himself will give expression to a limited number of delusions which he does not correlate. A few suggestive questions, however, will educe a mass of delusions, which when pieced together demonstrate the logical unconscious ideas that give rise to them. If such a patient be asked "What are your ideas?" he can give no reply. Ask him, however, if any one is mistreating him and you will start a train of thought in which one fancied insult leads to another or to delusions which do not represent mistreatment at all. On the other hand approach a patient with constructive delusions with the same question as to his ideas and he will produce a theory of the universe, often with a chronological account of how these ideas developed. He is insane in that his fancies do not reach an outlet in action being an end in themselves; but he is sane in so far as he keeps his ideas within the range of introspection and has not allowed them to become autonomous. The inferences from this to the laws of normal association are obvious.