Let us turn to the fragmentary nature of reproduced experience and discuss its significance. That previous experience can be reproduced only in fragments is the direct result of the selective power of attention, which asserts itself in both perception and ideation. Not every quality of a thing presented is equally interesting. A child having a watch takes interest mainly in the ticking and in the glitter of the golden case. Meeting a dog, he gives attention to the terrifying bark and the multiplicity of legs. Suppose now that the dog regularly occurred together with a special impression, perhaps a spoken word; then the recurring of this symbol will tend to reproduce in the child’s mind the image of the dog. But the pressure of many competing tendencies does not permit the reproduction of all the qualities of the dog which have become conscious on former meetings with this animal. Only an extract, so to speak, of these qualities is reproduced, and this is made up of those which were formerly especially interesting,--the bark and the legs. Another factor determining the selection of special qualities of a thing for reproduction is the frequency with which each quality reappears in things which are different in certain respects, but in other respects belong to the same class. The trees of a forest beside which I am walking have many individual differences. But certain features are common to all the trees. These common features reappear again and again, while each of the other features appears only now and then. The same can be said of various dogs met on the street, of various tones of a violin, and so on. If the perception of the trees is experienced together with a certain other percept which may serve as a symbol for the trees, for example the word _tree_, the association of the symbol with those regularly repeated qualities becomes firmly established, whereas the association with the other, more or less varying qualities, remains comparatively feeble. The result is that the symbol tends to reproduce almost exclusively the former qualities. These come to make up a separate group of images, a general idea. The laws of attention, practice, and memory, together with the simple uniformity of nature just mentioned, produce thus a peculiar result. They remove ideation from the accidents of external events in an incomparably higher degree than perception. They bring about ideas of the separate qualities of the things perceived, _abstractions_, and ideas of common features, _general ideas_. In many cases an idea is both an abstraction and a general idea. Examples of such ideas to which no equally simple concrete object corresponds, are the idea of a mere length, the color red, sight, a dog in general, a tree in general.