In order to get a clear idea of this, we must understand carefully the relation of the embryonic shield to the germinative area and the embryonic vesicle. This is done best by a comparison of the five stages which are shown in longitudinal section in Figs. 138–142. The embryonic shield (_c_), which at first projects very slightly over the surface of the germinative area, soon begins to rise higher above it, and to separate from the embryonic vesicle. At this point the embryonic shield, looked at from the dorsal surface, shows still the original simple sandal-shape (Figs. 133–135). We do not yet see any trace of articulation into head, neck, trunk, etc., or limbs. But the embryonic shield has increased greatly in thickness, especially in the anterior part. It now has the appearance of a thick, oval swelling, strongly curved over the surface of the germinative area. It begins to sever completely from the embryonic vesicle, with which it is connected at the ventral surface. As this severance proceeds, the back bends more and more; in proportion as the embryo grows the embryonic vesicle decreases, and at last it merely hangs as a small vesicle from the belly of the embryo (Fig. 142 _ds_). In consequence of the growth-movements which cause this severance, a groove-shaped depression is formed at the surface of the vesicle, the _limiting furrow,_ which surrounds the vesicle in the shape of a pit, and a circular mound or dam (Fig. 139 ks) is formed at the outside of this pit by the elevation of the contiguous parts of the germinal vesicle.
Fig.136. Transverse section of the embryonic disk of a chick at the end of the first day of incubation. Fig. 136—Transverse section of the embryonic disk of a chick at the end of the first day of incubation, magnified. The edges of the medullary plate (_m_), the medullary swellings (_w_), which separate the medullary from the horn-plate (_h_), are bending towards each other. At each side of the chorda (_ch_) the primitive segment plates (_u_) have separated from the lateral plates (_sp_). A gut-gland layer. (From _Remak._)
In order to understand clearly this important process, we may compare the embryo to a fortress with its surrounding rampart and trench. The ditch consists of the outer part of the germinative area, and comes to an end at the point where the area passes into the vesicle. The important fold of the middle germinal layer that brings about the formation of the body-cavity spreads beyond the borders of the embryo over the whole germinative area. At first this middle layer reaches as far as the germinative area; the whole of the rest of the embryonic vesicle consists in the beginning only of the two original limiting layers, the outer and inner germinal layers. Hence, as far as the germinative area extends the germinal layer splits into the two plates we have already recognised in it, the outer skin-fibre layer and the inner gut-fibre layer. These two plates diverge considerably, a clear fluid gathering between them (Fig. 140 _am_). The inner plate, the gut-fibre layer, remains on the inner layer of the embryonic vesicle (on the gut-gland layer). The outer plate, the skin-fibre layer, lies close on the outer layer of the germinative area, or the skin-sense layer, and separates together with this from the embryonic vesicle. From these two united outer plates is formed a continuous membrane. This is the circular mound that rises higher and higher round the whole embryo, and at last joins above it (Figs. 139–142 _am_). To return to our illustration of the fortress, we must imagine the circular rampart to be extraordinarily high and towering far above the fortress. Its edges bend over like the combs of an overhanging wall of rock that would enclose the fortress; they form a deep hollow, and at last join together above. In the end the fortress lies entirely within the hollow that has been formed by the growth of the edges of this large rampart.