In a study of the aortic cell-clusters in loggerhead-turtle embryos my attention was arrested by peculiar formations in the inferior vena cava at the level of fusion of the subcardinal veins of a specimen of the twelfth day of incubation. These formations consisted of encap sulated spheroidal masses of hemoblast-like cells, and long strings of hemoblasts attached to the endothelium and extending free into the lumen of the vessel. In the case of the encapsulated groups the capsule consists of an endothelioid membrane of greatly flattened cells. The inclosed cells are very similar to those of the naked cell-clusters of the aorta and the superior mesenteric artery. The structure is com parable to Emmel's figure 5, which represents a similar formation in the aorta of a 12 mm. pig embryo. In one instance the mass was con tinuous with an underlying loose mesenchyma which appeared to be differentiating into hemoblasts. I incline to interpret this structure in terms of my figure 5. If the invaginated area of endothelium had included a considerable portion of the subjacent vascularizing mesen chyma, then the mesenchyma might have outstripped the endothelium in the process of differentiating into hemoblasts, and so forced, through pressure, the peripheral endothelial cells to continue their further differentiation into definitive endothelium.
As to the endothelial strands of hemoblast-like cells, I incline to an interpretation in accord with the idea that slightly differentiated endothelium anywhere carries the capacity of producing hemoblasts. In a footnote (p. 407) Emmel describes similar strands in the aorta and the proximal portion of the left umbilical artery in the 12-mm. pig embryo which contained the encapsulated cluster, and in a second 12-mm. embryo, and suggests that they may be associated with the fusion of the two original dorsal aorta.
If one wishes to adhere to an interpretation of these proliferation products of the aortic endothelium in terms of a toxin, one might locate the source of the toxic substance in the mesonephros, where degenera tive processes are initiated in the anterior portion; but such a view is again contradicted by the observation that endothelial desquamation products are practically lacking in the glomerular capillaries, though occasionally present near the aortic mouth of the afferent arteriole.
The view that we are dealing with a normal hemogenic process related to a relatively undifferentiated condition of endothelium such as would seem to be requisite in the ventral portion of the abdominal aorta, to permit of the shifting of the celiac, superior mesenteric, and inferior mesenteric arteries by ahnost any reasonably conceivable process, seems to fit all the facts better than the idea of a toxic influence dependent upon a tissue degeneration in redundant atrophying blood-vessels.
In a second paper ("Concerning certain cellular elements in the ccelomic cavities and mesenchyma of the mammalian embryo," Amer. Jour. Anat., vol. 20, 1, 1916, pp. 73-125), Emmel describes the origin of "macrophages" from the pericardial and peritioneal mesothelium; he describes and illustrates also certain free cell-masses in the peri cardial cavity comparable with my illustration (fig. 21) in the mongoose embryo. The causal factor is again assumed to be a toxic substance, in this case liberated in part by degenerating erythrocytes in the ccelom. But it would seem quite as reasonable to regard this formation of macrophages in the embryo as a normal incident in the life of the embryo.
In conclusion, I believe that a careful consideration of all the facts relative to Emmel's observations and my own, upon which there is essential agreement, more amply justifies the conclusion that young, relatively undifferentiated endothelium and mesothelium may any where in the embryo normally produce hemoblasts (and macrophages), than that such activity demands the operation of a pathologic factor in the form of a dilute, slow-acting toxin.