Strümpell,66 and almost identically by Wundt,76 to the effect that the reaction of the mind to the attacking stimuli in sleep is determined by the formation of illusions. A sensory impression is recognised by us and correctly interpreted, i.e. it is classed with the memory group to which it belongs according to all previous experience, if the impression is strong, clear, and long enough, and if we have the necessary time at our disposal for this reflection. If these conditions are not fulfilled, we mistake the objects which give rise to the impression, and on its basis we form an illusion. "If one takes a walk in an open field and perceives indistinctly a distant object, it may happen that he will at first take it for a horse." On closer inspection the image of a cow resting may obtrude itself, and the presentation may finally resolve itself with certainty into a group of people sitting. The impressions which the mind receives during sleep through outer stimuli are of a similar indistinct nature; they give rise to illusions because the impression evokes a greater or lesser number of memory pictures through which the impression receives its psychic value. In which of the many spheres of memory to be taken into consideration the corresponding pictures are aroused, and which of the possible association connections thereby come into force, this, even according to Strümpell, remains indeterminable, and is left, as it were, to the caprice of the psychic life.
We may here take our choice. We may admit that the laws of the dream formation cannot really be traced any further, and therefore refrain from asking whether or not the interpretation of the illusion evoked by the sensory impression depends upon still other conditions; or we may suppose that the objective sensory stimulus encroaching upon sleep plays only a modest part as a dream source, and that other factors determine the choice of the memory picture to be evoked. Indeed, on carefully examining Maury's experimentally produced dreams, which I have purposely reported in detail, one is apt to think that the experiment really explains the origin of only one of the dream elements, and that the rest of the dream content appears in fact too independent, too much determined in detail, to be explained by the one demand, viz. that it must agree with the element experimentally introduced. Indeed