THE EMBRYOLOGY OF THE OPOSSUM 103 



so that it loses its other connections and becomes a tributary 

 of the external jugular. 



The capillary plexus which originally connected the aortae- 

 with the sinus terminalis has in stage 27 become organized 

 into two vitelline arteries which leave the aorta in the region 

 of the midgut, run caudad beneath the extra-embryonic coelom 

 and through the area vasculosa to the sinus. In advanced 

 specimens the two have fused into one in their extra- 

 embryonic portion. 



Miscellaneous details. There is a recognizable hind-gut 

 diverticulum or posterior intestinal portal in even the very 

 early specimens of stage 27, but no allantois. Section H, 

 figure 30, shows the posterior end of the body of a very early 

 specimen in sagittal section. It can be seen that a deep and 

 pointed proctodeum makes contact with an equally pointed 

 hind-gut. The contact is, of course, the cloacal membrane. 



Anterior to the hind-gut diverticulum another pocket in the 

 endoderm can be seen. This represents the vestige of a 

 notochordal canal. At this point there is no limiting mem- 

 brane on the dorsal side of the endoderm or on the ventral 

 side of the ectoderm, and endoderm, notochord, and Hensen's 

 node are continuous. Up until stage 26 there had been a 

 distinct primitive pit in Hensen's node, but in stage 27 this 

 pit closes. Selenka mentioned not having seen a neurenteric 

 canal, though he thought it likely that this was due to his 

 not having the appropriate stage. No neurenteric canal is 

 formed, but the endodermal pit which marks the site where it 

 nearly formed remains distinctly visible through stage 28 

 (fig. 37). 



In the earliest specimens of stage 27 (17141, 17142) the 

 neural tube is closed anteriorly only through the myelen- 

 cephalon the metencephalon, mesencephalon, and prosen- 

 cephalon being wide open (fig. 30, A). In specimen 17154 the 

 mesencephalon is closed but the metencephalon and prosen- 

 cephalon are still open (fig. 30, C). In the last six specimens 

 the metencephalon has closed and only a small neuropore re- 

 mains in the prosencephalon at the level of the optic stalks. 



