406 A HUMAN EMBRYO OF THE PRESOMITE PERIOD. 



the angioblastic strand into endothelial cells, which can be distinguished by the 

 shape of their nuclei and by their tendency to so arrange themselves as to form the 

 contour of the strand. The cytoplasm of the other components of the strand under- 

 goes liquefaction, either in the formation of large vacuoles or by resolution into a 

 very fine mesh, which also disappears. The appearances in this respect probably 

 represent the vacuolization phenomena demonstrated in the living chick by 

 Professor Sabin, whose preparations I have had the privilege of examining and whose 

 observations are reported elsewhere in this volume. In some cases we find a few 

 round nuclei persisting, either adherent to the endothelial wall or suspended between 

 the two endothelial walls by slender threads. These are doubtless to be regarded as 

 future blood-cells. Whereas the spaces in the vacuolated strands are first seen to 

 be incompletely closed off from the spaces of the surrounding stroma, further 

 differentiation of the endothelium gradually completes their boundaries, thus 

 forming completely closed tubes as established vessels. In other words, we are 

 dealing with the formation of multinucleated strands, some of the elements of 

 which become differentiated into endothelium, while the remainder is either com- 

 pletely liquefied or persists as blood-cells within the peripherally formed endothelial 

 elements. 



Figure 17 represents an angioblastic strand with its longitudinally arranged, 

 elongated nuclei. In some places there is a slight indication of cleavage of the cyto- 

 plasm of such a strand, but this is always incomplete. Around its margins it is more 

 or less continuous with the delicate trabeculse of the surrounding stroma. In figure 



18 a similar strand is shown extending as a lateral process from the margin of a 

 vessel that is farther advanced. Figure 19 illustrates the transition of a solid strand 

 into an endothelial tube. Here the nuclei are in active proliferation. It can be 

 plainly seen that the cells along the margin are elongating into typical endothelium, 

 as regards both the nuclei and the adjacent cytoplasm. The cytoplasm of the 

 central part of the strand shows enlarging vacuoles. At other points, such as the 

 right-hand end of the figure, instead of showing a simple, large space, thr cyto- 

 plasm becomes converted into a very degenerate mesh. As a result of these two 

 processes there is a general liquefaction, or conversion into plasma, of the central 

 portion of the angioblastic strand. Most of the nuclei seen in the strand in figure 



19 have either divided or are about to divide. However, it is evident that some of 

 them, together with their surrounding cytoplasm, must undergo degeneration. 



Figure 20 shows a strand in which there is left only the differentiated endothe- 

 lium. The incompletely closed lumen of this strand, as far as one can judge from 

 the sections, still seems to communicate with the spaces of the surrounding stroma. 

 In figure 21 we meet with a condition in which the endothelium forms a completely 

 closed tube. A transverse section of a similar vessel is shown in figure 22. In figure 

 21 a few trabecula3 still traverse the lumen connecting the opposite endothelial 

 walls. There can also be seen within the lumen an occasional large, round, nucleated 

 cell, showing a scant amount of cytoplasm, which apparently represents an embryonic 

 blood-cell. This is the most mature type of blood-vessel encountered in the spec- 

 imen. Up to this time there is apparently not a very active formation of blood-cells. 



