Hemopoiesis in the Mongoose Embryo. 301 



shifting of tlie larger stems (celiac artery, superior and inferior mesen- 

 teric arteries) from higher to lower ventral connecting segmental twigs. 

 This region would seem to contain a less highly differentiated type of 

 endothelium — that is, one of greater proliferative capacity, providing 

 thus for an unequal growth between the dorsal and ventral walls. 

 Herein may possibly reside the capacity for some of these endothelial 

 cells to differentiate into hemoblasts. If this interpretation is correct, 

 then the younger endothelium of the intra-embryonic vessels should 

 also be able to transform into hemoblasts. Such youngest type of 

 endothelium should be present in the numerous small blood-channels 

 in the anterior head region next the brain. 



ENDOTHELIAL HEMOGENESIS IN INTRAEMBRYONIC MESENCHYMA. 



Search for evidence in support of this hypothesis is abundantly 

 rewarded. Figure 13 represents a blood-channel which has just 

 formed out of the pericerebral mesenchj^ma. The lumen of the 

 channel still contains strands of dissolving mesenchyma. The four 

 cells present represent four different stages in the transformation of 

 a mesenchyme cell into a hemoblast, leading through an endothelial 

 cell. Cell a is still tyi3ically mesenchymal, but is clearly in process of 

 transformation into an endothelial cell, for two of its processes already 

 form portions of the wall continuous with that portion formed by the 

 typical endothelial cells b and c. The original endothelial cell now 

 represented by c and d apparently divided its nucleus, thus forming a 

 binucleated cell ; the distal portion of this cell underwent differentiation 

 into a typical hemoblast, while the proximal end represents a transition 

 stage between a typical endothelial cell and a true hemoblast. In this 

 single section of a small area, including four nuclei, the whole process 

 of mesenchymal and endothelial hemogenesis may be seen in abbre- 

 viated form; and the process is essentially identical with the manner 

 in which mesenchyma and endothelium function hemogenically in the 

 yolk-sac wall. 



Figure 14 is of a cross-section of a capillary vessel including a single 

 endothelial cell, and with a diameter just sufficient to accommodate a 

 single erythrocyte. Since this section is from the same vascularizing 

 area as that of figure 13, the most probable interpretation that sug- 

 gests itself is in terms of cells c and d of figure 13. If cell c, still con- 

 tinuous with the general mesenchyma, had differentiated into an 

 endothelial cell, and d into an erythrocyte, we would have exactly 

 the condition shown in figure 14. The endothelial wall here is still 

 continuous with the mesenchyma, as shown by the strands of proto- 

 plasm at the upper pole. If the cytoplasm inmiediately enveloping 

 the second nucleus of an originally binucleated cell differentiated into 

 hemoglobin-containing protoplasm and thus into an erythrocyte, while 

 the more peripheral cytoplasm remaining in functional association with 

 the second nucleus differentiated into endothelial protoplasm, the 



