It is not our task to examine the first four elements. In order clearly to understand the variety of perception, we have to deal with the last only. I once tried to explain this by means of the phenomenon of instantaneous photographs (cinematographs). If we examine one such representing an instant in some quick movement, we will assert that we never could have perceived it in the movement itself. This indicates that our vision is slower than that of the photographic apparatus, and hence, that we do not apprehend the smallest particular conditions, but that we each time unconsciously compound a group of the smallest conditions and construct in that way the so-called instantaneous impressions. If we are to compound a great series of instantaneous impressions in one galloping step, we must have condensed and compounded a number of them in order to get the image that we see with our eyes as instantaneous. We may therefore say that the least instantaneous image we ever see with our eyes contains many parts which only the photographic apparatus can grasp. Suppose we call these particular instances a, b, c, d, e, f, g, h, i, j, k, l, m; it is self-evident that the manner of their composition must vary with each individual. One man may compound his elements in groups of three: a, b, c,--d, e, f,--g, h, i, etc.; another may proceed in dyads: a, b,--c, d,--e, f,--g, h,--etc.; a third may have seen an unobservable instant later, but constructs his image like the first man: b, c, d,--l, m, n, etc.; a fourth works slowly and rather inaccurately, getting: a, c, d,--f, h, i,--etc. Such variations multiply, and when various observers of the same event describe it they do it according to their different characteristics. And the differences may be tremendous. Substitute numerals for letters and the thing becomes clear. The relative slowness of our apprehension of visual elements has the other consequence that we interpolate objects in the lacunæ of vision _according to our expectations_. The best example of this sort of thing would be the perception of assault and battery. When ten people in an inn see how A raises a beer glass against B’s head, five expect: “Now he’ll pound him,” and five others: “Now he’ll throw it.” If the glass has reached B’s head none of the ten observers have seen how it reached there, but the first five take their oath that A pounded B with the glass, and the other five that he threw it at B’s head. And all ten have really seen it, so firmly are they convinced of the correctness of their swift judgment of expectation. Now, before we treat the witness to some reproach like untruth, inattention, silliness, or something equally nice, _we_ had better consider whether his story is not true, and whether the difficulty might not really lie in the imperfection of our own sensory processes. This involves partly what Liebmann has called “anthropocentric vision,” i.e., seeing with man as the center of things.