Under mistakes of generalization the most important processes are those of arrangement, where the environment or accompanying circumstances exercise so determinative an influence that the inference is often made from them alone and without examination of the object in question. The Tanagra in the house of an art-connoisseur I take to be genuine without further examination; the golden watch in the pocket of a tramp to be stolen; a giant meteor, the skeleton of an iguana, a twisted-looking Nerva in the Royal Museum of Berlin, I take to be indubitably original, and indubitably imitations in the college museum of a small town. The same is true of events: I hear a child screeching in the house of the surly wife of the shoemaker so I do not doubt that she is spanking it; in the mountains I infer from certain whistles the presence of chamois, and a single long drawn tone that might be due to anything I declare to have come from an organ, if a church is near by. All such processes are founded upon experience, synthesis, and, if you like, prejudices. They will often lead to proper conclusions, but in many cases they will have the opposite effect. It is a frequently recurring fact that in such cases careful examination is most of all necessary, because people are so much inclined to depend upon “the first, always indubitably true impression.” The understanding has generalized simply and hastily, without seeking for justification. The only way of avoiding great damage is to extract the fact in itself from its environment and accompanying circumstance, and to study it without them. The environment is only a means of proof, but no proof, and only when the object or event has been validated in itself may we adduce one means of proof after another and modify our point of view accordingly. Not to do so, means always to land upon false inferences, and what is worse, to find it impossible upon the recognition of an error later on, to discover at what point it has occurred. By that time it has been buried too deep in the heap of our inferential system to be discoverable. The error of confusion Mill reduces especially to the unclear representation of _what_ proof is, i.e., to the ambiguity of words. We rarely meet such cases, but when we do, they occur after we have compounded concepts and have united rather carelessly some symbol with an object or an event which ought not to have been united, simply because we were mistaken about its importance. A warning example may be found in the inference which is made from the sentence given a criminal because of “identical motive.” The Petitio, the Ignorantia, etc., belong to this class. The purely logical mistakes or mistakes of syllogism do not enter into these considerations. Section 33. (k) Statistics of the Moral Situation.