There appears no valid reason why the endothelium of the yolk-sac vessels and that of the embryonic vessels should function differently; why in the case of the yolk-sac vessels the endothelium should be capable of metamorphosing normally into hemoblasts, while in the intra-embryonic vessels with young endothelium the stimulative factor to should have to be a pathologic one. If endothelial hemogenesis in the yolk-sac increased in activity coincident with the degenerative processes of the sac, Emmel's interpretation might be more confidently accepted; but this is precisely where it fails. Endo thelial hemogenesis is most active in the yolk-sac of the 10 to 12 mm. embryos when hemopoiesis is at its height in these vessels. Subse quently it decreases, and by the 25-mm. stage no separating endothelial cells can be found. Moreover, in the small vessels of the pericerebral region endothelial cells may occasionally separate to become intra vascular elements. In short, wherever only slightly differentiated endothelium appears, blood elements may originate. Again, if abnor mal conditions could explain endothelial proliferation and metamor phosis in embryos, then the experimental Fundulus embryos of Stockard, in which cessation of heart-pulsation produced stasis in the blood-vessels, and a consequent degeneration of some of the included cells, should show endothelial proliferative activity; but according to Stockard this endothelium is inactive.
Emmel lays much stress upon the connection between certain ventral aortic clusters and intra-arterial cell-masses. Vessels containing such masses are interpreted as "degenerating"; but the evidence that such are degenerating is unconvincing. In a 10-mm. pig embryo cut in transverse section I found a large cluster attached to one side of the mouth of the superior mesenteric artery. A similar condition prevails in the 5-mm. mongoose embryo. These vessels are not degenerating; if cut obliquely similar clusters might appear to be attached to an intra-arterial cell-mass filling the vessel. Emmel's section No. 6 (p. 419) is apparently obliquely cut. The "intra-arterial cell-mass" may be simply attached to one side of the mouth of the vessel. Such conditions occur in both the mongoose and pig embryo, where no sign of atrophy appears in the vessel itself. On the other hand, ventral branches which later disappear can be seen free of cell-clusters. In short, neither the degeneration nor the occlusion of all aortic branches associated with clusters is definitely proved.
Nevertheless, certain aortic branches undoubtedly do disappear in the caudal shifting of the ventral branches, and this process may be of the nature of a degeneration which may liberate a dilute and slowly acting toxic substance comparable to such .substances as stimulate the production of endothelial leucocytes in certain pathologic condi tions, e. g., the relatively slightly virulent toxins arising from typhoid and tubercle bacilli. Then this effect should either be felt throughout
the whole of the abdominal aorta, or there should be a progressively decreasing effect, as indicated by the abundance of the desquamating endothelial elements from the distal to the proximal (aortic) portion of these ventral rami. But neither of these conditions obtain. In the older embryos the clusters and desquamating cells are practically limited to the ventral wall; in the younger embryos (mongoose) where there are fewer atrophying ventral rami, desquamating cells can be found in the lateral and even the dorsal wall, and the ventral cell clusters are small. The younger the embryo the less differentiated the endothelium. Moreover, the clusters of the older embryos are near the mouths of the vessels, or farthest removed from the site of presumed intensest degeneration; and numerous vessels contain no clusters at all.
In the 7-mm. mongoose embryo several occluded ventral rami appear, comparable to Emmel's figure 7. The lumen of the vessel is completely filled with hemoblast-like cells similar to those of the aortic clusters, many of which show degeneration stages, principally a karyorrhexis. But the presence of these cells need not be interpreted as the result of the liberation of a toxin by the atrophying ramus. It seems more reasonable to suppose that the ramus contained a cluster of normal hemoblasts as the result of a normal hemogenic capacity on the part of the endothelium of the vessel, which cluster, in conse quence of the atrophy and coincident constriction of the vessel, came to occlude the lumen and ultimately to suffer a resultant degeneration. In other words, the occlusion of the lumen of the vessel and the kary orrhexis of the included hemoblasts are more probably secondary effects of the atrophy of the ramus than that the presence of the cells in the lumen is the result of a toxin formed by the degeneration of the vessels and the included cells and operating as a desquamating stimulant upon the endothelial cells of the vessel.
The most damaging countervailing evidence, however, to the inter pretation that the endothelial desquamation products are the result of the action of toxins produced by degenerating blood-vessels and blood-cells, consists in the presence (in the 10-mm. pig embryo) of endothelial cell-clusters of hemoblasts. deep within the superior mesenteric artery (that is, in the middle third of its extent), a level which shows no other signs of atrophy or degeneration and which suffers no subsequent change in a possible farther caudal progression of the vessel to its definitive point of attachment to the aorta. Also in the mongoose embryos and in turtle embryos, the endothelium of the superior mesenteric artery is especially active in liberating intravascular cellular elements.