Herbert Spencer pursues the same biological line. But he traces the impulse to song from the surplus of energy in living things. Nature, as we know, is prodigal, producing thousands of eggs or seeds where she expects only scores of insects or plants to grow. And so she gives to each living thing more energy than is needed to keep it alive and healthy--or at least more energy than is needed _all the time_. With this surplus of energy something must be done. And that something among men is art. When a wild rose-bush is put in rich soil it grows a double rose and a larger one. And so, when a man has a little more than enough for his material needs, he attempts to decorate and beautify his surroundings. To Spencer the typical example of nascent art is the child who dances out of pure joy of life. But others object to this, saying that this dancing of the child (and all similarly spontaneous art) is lacking in the very quality which makes art--namely, discipline and control. It is selection and control, they say, which makes any work of art more than a mere daub of color or conglomeration of sounds. Certainly the biological theories have not satisfactorily explained the second great element in artistic creation--the impulse to refinement. We have seen that once our typical folk-song was invented by the Roumanian peasant woman it could not stay as it was, but had to take new forms, changing continually, and, as we see it, usually for the better. What is it that makes men dissatisfied with a crude, formless melody and desirous of making it regular and organic? Perhaps it is no more than the impulse to play which children show when they arrange pebbles in new and strange shapes on the sand or savages when they daub their faces with new colors of war-paint. Again the surplus of energy? Perhaps. Perhaps when people have become familiar with their primitive melody they notice a peculiar turn of the phrase and take pleasure in seeing what they can _do_ with it in the way of repeating it and inventing contrasting phrases. At any rate, there it is, this impulse to refinement, more mystery for the theorist and more entertainment for the music lover. On the whole, these theories probably err in trying to simplify the case too much. More likely music has not one simple _cause_, but many interacting _causes_. Like every individual man whose ancestors are a multitude, song must be the result of many, many causes. VI