FROM LAYING TO FORMATION OF FIRST SOMITE 89 
it is split off from the germ-wall (His and others). One thing 
is perfectly clear, viz., that the mesoderm of the opaque area 
arises in continuity with the primitive streak mesoderm; the 
second view would therefore be better expressed, as Riickert 
states it, that the primitive streak mesoderm grows in the region 
of the area opaca at the expense of elements of the germinal wall. 
If the cells of the primitive streak mesoblast be compared 
with the cells of the forming blood-islands a sharp contrast is 
observed; the mesoblast cells of the area pellucida are devoid of 
yolk-granules; young blood-islands on the other hand contain 
yolk-granules of precisely the same character as those of the 
germ-wall (Fig. 41), which must have been derived from the latter. 
If the origin of the blood-islands be carefully traced, they are 
found to be rooted in the protoplasm of the germ-wall; and prior 
to the appearance of the blood-islands proper, protoplasm and 
nuclei of the germ-wall aggregate superficially in a manner that 
appears to foreshadow the blood-islands. Therefore, either the 
blood-islands are derived from the celis of the germ-wall, or 
cells of the mesoderm growing over the germ-wall burrow into 
the latter, engulf yolk-spheres, and reappear in masses as blood- 
islands. We shall not attempt to decide between these pos- 
sibilities. 
Another question concerns the origin of the layer of ecelomic 
mesoblast that overlies the blood-islands: is it derived from the 
primitive streak mesoblast, or is it split off from the blood-islands ? 
When the latter first appear, in the periphery of the vascular 
area at least, there is no ccelomic mesoblast above them. It 
appears later, at first not as a coherent layer, but as scattered 
cells that rapidly unite to form a layer. In many places the 
microscopical appearances indicate strongly that the cells are 
split off from the surface of the blood-islands; but, as they are 
usually not far from the edge of the advancing ccelomic meso- 
blast, it may be that they are derived from the latter. Riickert 
states, however, that, in the case of some isolated blood-islands 
behind the embryo, a layer of mesoblast is formed over them 
while they are still isolated. This would render the derivation 
from the blood-islands probable in such cases. It is possible, 
therefore, that the caelomic mesoblast grows partly, at least, at 
the expense of the superficial cells of blood-islands. 
As rapidly as they are formed the various blood-islands con- 
