This is an example of a random train of ideas. It is that loose “thinking” referred to in our first chapter. But even this is made possible only by the connection of ideas in our mind at some previous period. No thought can enter our minds unless it is associated in some way with the previous thought. Psychologists have traditionally classified associations into four kinds: association by succession, by contiguity, by similarity and by contrast. The example just given involves all four. Association by succession means that when two ideas or impressions of objects have entered the mind in succession, the second is likely to be suggested whenever the first is thought of. A tune consists in a succession of notes, and when the first notes are brought to mind, as by a passing band, the rest will follow—sometimes in spite of ourselves. Association by contiguity means that when two objects or ideas have been in consciousness together, one is always likely to suggest the other thereafter. This was the case with the music and the dance, or the music and the two-step. Association by similarity occurs when two ideas resemble each other in some particular. They need not have occurred together at any past time, nor after each other. The fact that they have a common element suffices to bring up one idea when the other is in mind: thus the two-step suggested the one-step. Association by contrast needs no explanation. It is exemplified when the idea of present-day dancing brings up the idea of distant dancing. Any attempt to show _why_ the mind acts in this way, any explanation of the way in which the different kinds of association are made possible, would bring us into physiological psychology, would involve a study of the brain and the nervous system. For our purposes it is sufficient to keep in mind that such associations do take place. Without them no idea can occur. Without them thought is impossible. The bearing of all this on concentration has yet to be made plain. We must remember that every idea has more than one associate; in fact that each idea generally has a cluster of possible associates. Instead of suggesting the minuet, the one-step may have made the fox trot or the three-step occur to the young lady. It may have made her think of a young man with whom she danced it, or the trouble she had in learning it. Each of these suggestions, in turn, would also have potential connections with a cluster of ideas. When we are thinking at random—when we are day dreaming, as in the example given—the strongest association, or the first to be aroused, is the one we dwell upon. But when we are thinking with a purpose, in a word, when we are reasoning, we reject all associations which have no bearing on our purpose, and select only those which serve it.