292 RICHARD E. SCAMMON 



Similar changes are seen in earlier stages in the main median 

 lobe of the liver and in the process of the cystic lobe which ex- 

 tends into the yolk stalk coelome. 



In Mustelus and Torpedo the early development of the hepatic 

 sinusoids is quite different from that seen in Acanthias. In 

 these forms the omphalo-mesenteric veins become much enlarged 

 at an early stage and lie both dorso-medial and ventro-lateral 

 to the hepatic pouches instead of only medial to them in Acanthias. 

 As a result the hepatic tubules are in contact with the walls of 

 large vascular chambers from the time of their first appearance 

 and as they grow they can only do §o by pushing the endothelial 

 walls of the vessels before them. The tubules, as has been re- 

 marked, are slender and elongated and unite by anastomoses 

 soon after their formation. Thus there is no stage in these forms 

 corresponding to the early one in Acanthias where the sinusoids 

 are small cleft-like spaces. The proportion of the volume of the 

 liver formed by sinusoids is very high at first but is steadily 

 reduced as development proceeds. Figure 16 of two transverse 

 sections of Mustelus embryos 16.7 mm. and 22.5 mm. in length 

 respectively illustrates these points, as do also the thick frontal 

 sections shown in figures 44 and 45. The larger vasculartrunks of 

 the liver in Mustelus and Torpedo are mapped out by the growth 

 of the hepatic cylinder network between them. As in Acanthias 

 the main liver veins are marked out early in the development 

 of the organ, and are not formed by the fusion of smaller vascular 

 spaces. 



In both the forms under discussion and in Acanthias the 

 hepatic sinusoids are formed by the intercresence of the hepatic 

 cylinders and the omphalo-mesenteric veins. In Mustelus and 

 Torpedo this intercresence is due to the invasion of the space 

 occupied by the vein. In Acanthias, in early stages, the inter- 

 cresence is due to the penetration of the venous sprouts about 

 the tubules which are already established. In later stages in 

 Acanthias the cylinders increase by growing into venous spaces 

 and the end result in either case is the same. Thus both the 

 methods of sinusoid development postulated by Minot ('00) and 

 by Lewis ('04) are found in these forms. 



