210 Venous System of Didelphys Marsupialis (1) 
arteries between the external and internal iliac veins, either in the form 
of vessels which lie in the median line, or of those which are more lat- 
erally situated and which have been described above as the ventral por- 
tions of the lateral rings. Numerous examples were met with by the 
writer in which the postcava was formed in the adult in exactly the same 
manner as that represented by Fig. 14 (see Figs. 6, 7 and 8, Plate II, 
Part I) ; while the two variations represented by Figs. 9 and 10 on the 
same plate, are easily explained on the ground that the internal iliac 
veins have given up one of their connections with the external iliacs. 
From what has already been said regarding the pouch young there 
can be no doubt that the establishment of a particular type of postcava in 
the adult depends upon the manner as well as the extent to which certain 
vessels which he dorsal and ventral to the iliac arteries are affected by 
atrophy. There can also be no doubt but that, in all cases, the vessels 
which lie dorsal to these arteries are derived from the postcardinal veins. 
There is, however, some doubt in the case of certain vessels which extend 
between the iliac veins, ventral to the arteries. A doubt may exist whether 
all of these ventrally situated vessels have been derived from the ventral 
portions of embryonic circumarterial venous rings; or, whether they may 
not, in some cases, be new formations which have been developed in the 
pouch young, independently of these rings. 
It is probable that in such a variable venous system as that of Didel- 
phys any venous channel which might be established in the embryo or 
pouch young between the internal and external iliac veins could, in cer- 
tain circumstances, be retained in the adult as a functional channel. I 
am, therefore, inclined to believe that in addition to the embryonie cir- 
cumarterial venous rings, other venous elements may occasionally enter 
into the formation of the vessels which form an anastomosis between 
the iliac veins ventral to the common iliac arteries. 
In all probability, the median ventral vessel in certain cases under 
Type I, as in Figs. 1 and 2 (Plate I, Part I) is formed as the result of 
the persistence of a single median vessel similar to that met with in 
the 15 mm. embryo (Text Fig. 17). It is a difficult matter to determine, 
however, whether both of the anastomoses between the iliac veins as in 
Fig. 3 (Plate I, Part I) have been derived from the ventral portions of 
the lateral rings, or, whether such is the case only with the anastomosis 
on the left side. It is also a question, on account of their connection with 
the Vy. pudendovesicales, whether the two veins which unite the iliac 
veins ventral to the arteries in Fig. 14 (Plate III, Part I) are not de- 
rived from the ventral portions of the lateral rings rather than from the 
ventral portions of the embryonic circumarterial venous rings. This and 
