482 
vein does not break up into capillaries in the liver to form a portal 
system, although it runs over the surface of, and even through the 
liver tissue. This is evidently an extremely primitive character, for 
the same relation is found in the early embryonic development of the 
same vessel in Petromyzon and all higher vertebrates.” ‘It receives. 
branches from the posterior lobe of the liver, and becomes the posterior 
hepatic vein.” It seems reasonable to conclude that the sub-intestinal 
vein of Myxinoids is the persisting sub-intestinal + vitelline vein of 
the embryo. It has, however, acquired a connection with the posterior 
lobe of the liver, and has degenerated posteriorly, although the extent: 
to which the latter happens varies greatly in different individuals. 
The feature of greatest interest in the Myxinoid vascular system: 
is the extensive series of spaces now known as the veno-lymphatics. 
There are grounds for believing that these spaces are only enlarged 
veins, and are not comparable with the true lymphatics of higher 
animals. Such, in fact, are the conclusions to be drawn from the 
recent careful work of MoZEJKo and ALLEN. These spaces always. 
contain blood—in Myxine, as far as my experience goes, in con- 
siderable quantity, and they are said to develop as true venous channels. 
The question therefore arises, how does the blood enter and leave 
them? As regards the former point, the only statements available 
are those contained in my papers published in 1905 and 1912, in 
which it is shown that blood may enter the peribranchial sinuses. 
from the afferent branchial arteries. The efferent or discharging 
vessels of the veno-lymphatics have been described in detail by MozEsKo- 
in the Lamprey, but not in the Myxinoids except in the case of the 
caudal circulation. I have therefore searched for these channels with. 
the following results: a) in my large series of sections the last right. 
peribranchial sinus opens by a valved aperture into the left anterior 
cardinal, but in two dissections both the last peribranchial sinuses 
opened into the anterior cardinal anastomosis; b) the lymphatic spaces 
of the club muscle open posteriorly by a valved aperture into the 
inferior jugular vein at the point where the latter assumes the median. 
position, i. e. immediately behind the club muscle. This connection. 
was found both in injected material and in serial sections. A forward 
injection of the inferior jugular vein fails to reach the sinuses of the: 
club muscle owing to the valve. Therefore the contents of the sinuses 
must drain posteriorly into the inferior jugular vein. It is interesting: 
to note that in the Lamprey Mozrsxo finds a similar communication. 
