In the cleavage stages the sea urchin's egg can be divided into five transverse layers, an1, an2, veg1, veg2, and the micromeres (Fig. I). The ectoderm of the pluteus larva is derived from an1+ an2+ veg1, while veg2 gives rise to the secondary mesenchyme, the coelom, and the endoderm. The micromere material migrates into the blastocoele before gastrulation, forming the primary mesenchyme, which produces the skeletal rods. The position uf the skeleton-forming cells and of the rods is determined by the ectoderm.

The factors determining the cleavage type of the 16-cell stage (8 meso-, 4 macro-, and 4 micromeres) seem to be (I) progressive changes in the cytoplasm, causing spindles formed a certain time after fertilization to lie in a certain direction, (2) the presence in the vegetative part of the egg of a region of micromere-forming material, and (3) the activation of that material a certain time after fertilization. This leads to partial cleavage of isolated blastomeres, and to whole, intermediate or partial cleavage of fragments of undivided eggs, depending upon the time and plane of isolation. Whole eggs may also show partial cleavage. Differentiation is independent of the type of cleavage which the egg or fragment has undergone.

The polarity (animal-vegetative axis) of the egg is fairly stable, since it is not altered by centrifuging or by moderate stretching, and it is more or less retained in fragments. On the other hand, the polarity can be changed both by a greater degree of stretching and by placing animal and vegetative material in atypical relationship to one another. A new axis may then be induced, and the whole polarity may sometimes be reversed. A reversal can be brought about both by vegetative and by animal material. The dorso-ventral axis is less stable, as it adjusts itself in accordance with the direction of stretching (perpendicular to the egg axis), not only to a considerable, but also to a moderate degree of stretching. After considerable stretching or constriction, both of which involve partial physiological isolation, as well as on complete isolation, the dorso-ventral axis is spontaneously reversed in the dorsal half. In right and left halves the dorso-ventral axis is maintained, and the larvae are more or less defective on thecut side. In starfish larvae a similar reversal of the right-left axis may occur in right halves.

After isolation animal material will form a considerably enlarged apical tuft and later develop only into ciliated cylinder epithelium. In veg., weg., and the micro-meres, we find the faculty of checking the enlargement of the apical tuft and of causing the formation of stomodaeum, ciliated band, and pavement epithelium out of the animal material. Moreover, veg2 has the faculty of forming endoderm and skeletal cells, and under certain conditions also ectoderm. An endodermization of presumptive ectoderm can also be brought about by veg2, but this power of induction is much stronger in the micromeres. To explain the conditions in the sea urchin's egg we assume an animal and a vegetative gradient, both reaching the opposite pole and progressively diminishing. The animal and the vegetative qualities or forces have to interact in order to bring about normal differentiations, e.g. vegetative influences are necessary for the formation of ciliated band and stomodaeum, animal ones for gastrulation and skeleton formation, and so on. The differentiation depends—within wide limits—upon the relative amounts of animal or vegetative material present. We find an endodermization both when the vegetative material is relatively increased (vegetative halves) and decreased (8 + 2 + 2), provided the decrease is not too extreme. In this last case (8 + 1/2 + 0) the vegetative properties may be suppressed. But thanks to a reconcentration at the poles of fragments, the animal and vegetative forces are often able to express themselves more strongly than would be expected from the prospective significance of the material or from the amounts present (apical tuft and mouth in vegetative halves, ectoderm in isolated veg2, as animal differentiations; skeletal cells in 8 + 4 + 0, 0 + 4 + 0, o +veg2+ o, as vegetative differentiations, and so on).

Implanting micromeres of one species into the blastocoele of another, von Ubisch has studied the formation of the skeleton in such germ-layer chimaeras (see the article by von Ubisch in Biological Reviews, vol. 14, 1939, p. 88). The roles of nucleus and cytoplasm in the formation of the skeleton have also been studied in heterosperm merogones (larvae with nucleus of one species and cytoplasm of another). A species character was found to follow the nucleus. Both in the germ-layer chimaeras and in the merogones the conditions are too complicated to allow a conclusion as to any possible role of the cytoplasm.