are made of literally hundreds or perhaps thousands of separate memories of sense impressions. We receive sense impressions at the rate of hundreds every day, and very many of these, perhaps all of them, are registered as memories and are fitted into their proper associations. The nervous machinery for carrying on different processes of association consists of nerve cells of the brain. These are all of the kind known as connecting cells. It will be remembered that the connecting cells may have many nerve junctions leading into them and they in turn are much branched so that many nerve junctions can lead out from them. The connecting cells of the brain are richer in these respects than those in any other part of the body, and the cells of the human brain than the brain of any of the lower animals. We suppose that the associations are formed by the passage of nervous disturbances from the arrival platforms where the sense memories are registered over various connecting nerve cells to a common meeting point in some cell or in some group of cells, where the associated memories are all brought together into a single memory complex. It is supposed that the large areas of the brain which are not taken up either by the arrival platforms or by the motor areas are the regions in which these associations take place. There is one particular group of associations that are so interesting as to call for special mention. These are the associations concerned with language. We are so in the habit of using language that we are likely never to have thought of it in its real meaning as a part of our mental activity. We know that we have two kinds of language, spoken and written, and that spoken language consists of certain sounds, and written language of certain visual symbols. Two of the distance senses then are concerned, sight for written language, hearing for spoken, and language itself consists simply of sense impressions coming in through one or the other of these sensory channels. The important feature of language is that mankind has selected arbitrarily certain sounds or certain written symbols to stand for particular things. When a child is learning to understand what is said, what he is really doing is fitting a particular set of arbitrary sense impressions into their proper places in his associations. We can illustrate this by the same example that was used a moment ago. The child becomes thoroughly familiar with its mother, so far as sense impressions are concerned that come directly from her. In course of time it adds to the associations thus formed an additional one made up of the sound of the word “mother.” Of course there is no particular reason why this sound should have that meaning rather than any other. The proof is that different languages have different sounds which stand for the same thing. After this sound has once been selected and learned, it becomes as much a part of the idea of