107. But here it is needful to recall a consideration sometimes disregarded, especially by those who speak of Differentiation as if it were some magical Formative Principle, quite independent of the state of the organized substance which is formed. There is a luminous conception--first announced by Goethe, and subsequently developed by Milne Edwards--which regards the organism as increasing in power and complexity by a physiological “division of labor,” very similar to that division of employments which characterizes the developed social organism. But the metaphor has sometimes been misleading; it has been interpreted as indicating that Function creates Organ (see PROBLEM I. § 88), and as if Differentiation itself were something more than the expression of the changes resulting from the introduction of different elements. In the Social Organism a “division of labor” presupposes that laborers with their labor-materials are already existing; the change is one of rearrangement: instead of each laborer employing his skill in doing many kinds of work, he restricts it to one kind, which he is then able to do with less loss of time and power. Thus is social power multiplied without increase of population, and the social organism becomes more complex by the differentiation of its organs. It is not precisely thus with the Animal Organism during its evolution. Indeed to suppose that the differentiation of the germinal membrane into special tissues and organs takes place by any such division of employments, is to fall into the ancient error of assuming the organism to exist _preformed_ in the ovum. The unequivocal teaching of Epigenesis is that each part is produced out of the elements furnished by previous parts; and for every differentiation there must be a difference in composition, structure, or texture--the first condition being more important than the second, the second more important than the third. The word protoplasm has almost as wide a generality as the word animal, and is often used in forgetfulness of its specific values: the protoplasm of a nerve-cell is not the _same_ as that of a blood-cell, a muscle-cell, or a connective-tissue cell, any more than a bee is a butterfly, or a prawn a lobster. No sooner has the specific character been acquired, no sooner is one organite formed by differentiation, than there is an absolute barrier against any transformation of it into any other kind of organite. The nerve-cell, muscle-cell, and epithelial cell have a common starting-point, and a community of substance; but the one can no more be transformed into the other than a mollusc can be transformed into a crustacean. In the homogeneous cellular mass which subsequently becomes the “vertebral plates,” a group of cells is very early differentiated: this is the rudimentary spinal ganglion, which becomes enveloped in a membrane, and then pursues a widely different course from that of the other