Only in rare cases will the resemblance between a fact and its theoretical conception extend _further_ than we ourselves postulate. Then the theoretical conception may lead to the discovery of _new_ facts, of which conical refraction, circular polarisation, and Hertz’s electric waves furnish examples that stand in opposition to those given above. But as a general rule we have every reason to distinguish sharply between our theoretical conceptions of phenomena and that which we observe. The former must be regarded merely as auxiliary instruments that have been created for a _definite_ purpose and which possess permanent value only with respect to that purpose. No one will seriously imagine for a moment that a real circle with angles and sines actually performs functions in the refraction of light. Every one, on the contrary, regards the formula sinα/sinβ = _n_ as a kind of geometrical model that _imitates in form_ the refraction of light and _takes its place_ in our mind. In this sense, I take it, all the theoretical conceptions of physics—caloric, electricity, light-waves, molecules, atoms, and energy—must be regarded as mere helps or expedients to facilitate our viewing things. Even within the domain of physics itself the greatest care must be exercised in transferring theories from one department to another, and above all more instruction is not to be expected from a theory than from the facts themselves. But instances were not lacking that demonstrated to me, how much greater the confusion was which was produced by the direct transference of theories, methods, and inquiries that were legitimate in physics, into the field of psychology. Allow me to illustrate this by a few examples. A physicist observes an image on the retina of an excised eye, notices that it is turned upside down with respect to the objects imaged, and puts to himself very naturally the question, How does a luminous point situated _at the top_ come to be reflected on the retina _at the bottom_? He answers this question by the aid of dioptrical studies. If, now, this question, which is perfectly legitimate in the province of physics, be transferred to the domain of psychology, only obscurity will be produced. The question why we see the _inverted_ retina-image _upright_, has no meaning as a psychological problem. The light-sensations of the separate spots of the retina are connected with sensations of locality from the very beginning, and we _name_ the places that correspond to the parts down, _up_. Such a question cannot present itself to the perceiving subject.