286 BICHARD E. SCAMMON 



tance a secondary hepatic duct derived from one of these tubule 

 clusters. As a rule the tubule ridges of the lateral surfaces of 

 the hepatic pouches are completely formed and are beginning 

 to break up into rows of individual anlagen before these blood 

 vessels invade their vicinity. Different embryos, however, show 

 some variation in this respect. The vessels pass backward 

 between these rows of tubules and form a vascular plexus inter- 

 rupted by a few small lacunae. After this process is quite well 

 under way, the main vascular trunk of the left omphalo-mesenteric 

 vein is interrupted by the growth of the pars hepatica medialis 

 and as a result vascular sprouts extend out from the posterior 

 part of this vessel and its fellow of the opposite side. These 

 sprouts join with the plexuses of the hepatic veins and there is 

 thus produced a set of sinusoidal vessels connecting the remains 

 of the omphalo-mesenteric (hepatic-portal) venous system with 

 the true hepatic veins. The extension of this network is gradual, 

 and is not completely established until the tubules are well differ- 

 entiated and anastomosis between the tubules begins to take 

 place. The establishment of the complete sinuosidal circulation 

 and of tubule anastomosis is practically simultaneous. In 

 Acanthias, as in the pig according to Hilton ('03) and in reptiles 

 (Hammer '93), there is no reason to believe that the blood vessels 

 have anything to do with the formation of the earliest hepatic 

 tubules. Certainly there is no breaking up of a solid mass of 

 liver cells as described in the frog by Shore ('93) or any indentation 

 of the wall of the lateral liver pouches as has been seen in Torpedo 

 by Choronshitzky ('00) . The vessels last of all penetrate among 

 the tubules of the ventral surface of lateral hepatic pouches. 

 Figure 12 is a frontal section through the tubules of this region 

 of an embryo 15 mm. in length (H.E.C. 228). It will be seen 

 that the tubules are separated only by scattered mesenchymal 

 cells with the exception of a single vascular sprig. 



Minot ('00) has remarked that in the liver, and in the pro- 

 nephros and mesenephros as well, the sinusoids which at first 

 may be small, increase in size until they reach a certain maximum 

 and then decrease. To secure some idea of the extent of the 

 changes which take place in this process, the area in cross-section 



