Cases like the foregoing show the organism affected during its process of transformation by large elemental influences, and the response to these influences is so familiar that often it does not surprise us. We veil the real mystery of the process by talking of the chemical and other physical properties of protoplasm which render this response possible. But when we come to the protective mimicry of stinging insects by stingless ones, of leaves by butterflies, and so forth, these physical explanations manifestly fail us. The explanation which assumes the building up of these extraordinary resemblances bit by bit, through natural selection working upon a multitude of fortuitous variations, fails us as completely. It would be difficult to accept it if only a single species of insect showed these mimetic markings. The unlikelihood of their production by mere chance in the case not of one but of hundreds of species of butterflies, flies, and caterpillars is stupendous, and defies all calculation. It must, we repeat, always be borne in mind that, if chance variations are all we can postulate, these variations must at first be confined to one or few individuals, and that the influence of intercrossing would always be at work to obliterate individual peculiarities before they could develop to the point of affording any protection worth mentioning. We are bound, therefore, so far as I can see, to conclude, first, that these mimetic markings originate not in individuals but in the species as a whole, and are an expression of the communal life of the species; secondly, that they are a real and direct response to the external conditions of danger from attacks of birds, etc., and of protection afforded by deceiving these foes through mimicry of something which they do not care to attack. They can only originate in the dominants of the reproductive cells, and there, where undoubtedly forces and affinities of which we have no conception are ever at work, the initial changes take place. These changes, no doubt, take place by forming new combinations or modifications of existing dominants. The directive force must have something to work on. It does not follow that because some things are possible to it therefore all things are. It is not to be expected, for instance, that human beings, although it would be a great advantage to them to fly, could ever develop wings, like the conventional angels of mediæval art, for that would violate the essential character of the archetypal form. It is true, however, that life is ultimately responsible for the material with which it works as well as the directive agency that breathes through it. This point is of importance and must be made perfectly clear. The view of cosmic action here put forward does not contemplate ‘interventions’ in the order of nature from a source outside it. There never was a moment when, if _law_ prevailed, one result would take place, while