Probably, even at the threshold of metazoan life, such variations did not affect only individual cells, but rather groups of cells. In other words, the differentiation was at once and primarily a tissue-differentiation. What do we know, however, about the primitive tissue-differentiation of the earliest metazoa? Hardly anything. We may fairly suppose that the first marked difference to appear was that between the outside and the inside. In the formation of an embryo this is the first differentiation we notice. From the beginning of segmentation or, in any case, very early, the outer-layer cells become marked off from the inner-layer cells. The next step was, perhaps, the formation of the mid-layer between the outer and inner. But how further differentiations were effected we really do not know, though we may guess a little. This, perhaps, we may fairly surmise--that fresh differentiations presupposed previous differentiations, and formed the basis of yet further differentiations. Thus calcified cartilage presupposes cartilage, and leads up to the formation of true bone. In all this, however, we are very much in the dark. We can watch, always with fresh wonder, the genesis of tissues in the development of the embryo; but we do not at present know much of the mode of their primitive genesis in the early days of organic evolution: how can we, then, pretend to understand their origins? If we speculate at all on the matter, we are led to the view that the variations must be primarily due to the differential incidence of mechanical stresses and physical or chemical influences. It may be admitted that this is little more than saying that they are due to some physical cause. Still, this at least may be taken as certain for what it is worth--that the primitive tissue-differentiations are due to physical or chemical influences, direct or indirect, on the protoplasm of the cell. Here is one mode of the origin of variations.