J. L. Bremer 4°73 
venosus in front and the ventral venous plexus behind; but the two 
models would come into actual contact only at the points marked and 
already noted. As both the liver and the venous plexuses continue to 
grow, they will come into contact over more and more of their extent, 
until finally the result of small, anastomosing channels, the endothelium 
of which is in close contact with the liver cords, will be produced. The 
method by which this result is attained is, however, in this embryo dif- 
ferent from that usually described; in the one case, the cords of liver 
cells push against the walls of a pre-existing large venous channel, and 
Fig. 14. Model of liver, from caudal end. X 160. 
by invaginating these walls, make finger-like trabeculee within this large 
space, yet separated from the cavity by the endothelium which invests 
them. ‘These cords then anastomose until the cavity of the vein is sub- 
divided into many smaller channels, the sinusoids, all lined by the origi- 
nal endothelium, and having for walls, beside this endothelium, the cords 
of liver cells. In this embryo, on the other hand, the cords of liver cells 
and the small branches from the vitelline veins are both pushing into 
the same mesenchyma, both growing rapidly, until, the mesenchyma re- 
maining small in amount, the whole enlarged mass will be made up 
practically of the veins and the liver cells, which by their continued 
growth must come into contact, the veins filling practically all the spaces 
