commission smelt a disgusting odor while a coffin was being exhumed only to discover that it was empty. If the coffin, for one reason or another, had not been opened all those present would have taken oath that they had an indubitable perception although the latter was only inferred from its precedent condition. Exner[154] cites the excellent example in which a mother becomes frightened while her child cries, not because the cry as such sounds so terrible as because of its combination with the consciousness that it comes from her own child and that something might have happened to it. It is asserted, and I think rightly, that verbal associations have a considerable share in such cases. As Stricker[155] expresses it, the form of any conceptual complex whatever, brings out its appropriate word. If we see the _thing_ watch, we get the _word_ watch. If we see a man with a definite symptom of consumption the word tuberculosis occurs at once. The last example is rather more significant because when the whole complex appears mistakes are more remote than when merely one or another “safe” symptom permits the appearance of the word in question. What is safe to one mind need not be so to another, and the notion as to the certainty of any symptom changes with time and place and person. Mistakes are especially possible when people are so certain of their “safe” symptoms that they do not examine how they inferred from them. This inference, however, is directly related to the appearance of the word. Return to the example mentioned above, and suppose that A has discovered a “safe” symptom of consumption in B and the word tuberculosis occurs to him. But the occurrence does not leave him with the word merely, there is a direct inference “B has tuberculosis.” We never begin anything with the word alone, we attach it immediately to some fact and in the present case it has become, as usual, a judgment. The thought-movement of him who has heard this judgment, however, turns backward and he supposes that the judge has had a long series of sense-perceptions from which he has derived his inference. And in fact he has had only one perception, the reliability of which is often questionable.