so as finally to evolve conduct which differs very widely with different circumstances and different characters. Even though it be true, as determinists believe (and I reckon myself among them), that such conduct is the necessary result of a given character and given circumstances--or, if you will, of a particular set of nervous structures and a particular set of external stimuli--yet we all know that it is capable of varying so indefinitely, owing to the complexity of the structures, as to be practically incalculable. But it is not so with the butterfly. His whole life is cut out for him beforehand; his nervous connections are so simple, and correspond so directly with external stimuli, that we can almost predict with certainty what line of action he will pursue under any given circumstances. He is, as it were, but a piece of half-conscious mechanism, answering immediately to impulses from without, just as the thermometer answers to variations of temperature, and as the telegraphic indicator answers to each making and breaking of the electric current. In early life the future butterfly emerges from the egg as a caterpillar. At once his many legs begin to move, and the caterpillar moves forward by their motion. But the mechanism which set them moving was the nervous system, with its ganglia working the separate legs of each segment. This movement is probably quite as automatic as the act of sucking in the new-born infant. The caterpillar walks, it knows not why, but simply because it has to walk. When it reaches a fit place for feeding, which differs according to the nature of the particular larva, it feeds automatically. Certain special external stimulants of sight, smell, or touch set up the appropriate actions in the mandibles, just as contact of the lips with an external body sets up sucking in the infant. All these movements depend upon what we call instinct--that is to say, organic habits registered in the nervous system of the race. They have arisen by natural selection alone, because those insects which duly performed them survived, and those which did not duly perform them died out. After a considerable span of life spent in feeding and walking about in search of more food, the caterpillar one day found itself compelled by an inner monitor to alter its habits. Why, it knew not; but, just as a tired child sinks to sleep, the gorged and full-fed caterpillar sank peacefully into a dormant state. Then its tissues melted one by one into a kind of organic pap, and its outer skin hardened into a chrysalis. Within that solid case new limbs and organs began to grow by hereditary impulses. At the same time the form of the nervous system altered, to suit the higher and freer life for which the insect was unconsciously preparing itself. Fewer and smaller ganglia now appeared in the tail segments (since no legs would any longer be needed there), while more important ones sprang up to govern the motions of the four wings.