DEVELOPMENT OF THE BLOOD VESSELS IN THE FROG. 199 



and slung to the dorsal wall of the pericardium by the mesocardial 

 septum. Its walls consist of two layers of cells : an outer thick 

 la} 7 er derived from the splanchnic mesoblast, and an inner layer of 

 flattened cells formed by the cardiac endothelium, the origin of which 

 has been discussed above. The heart is, in preserved specimens, 

 sometimes filled with a coagulum, but no blood corpuscles are yet 

 present in it, unless a few isolated cells, very similar in appearance 

 to the endothelial cells, though rather more spherical in shape, and 

 which in some specimens lie in the cavity of the heart, are to be 

 regarded as corpuscles. 



Posteriorly the heart is formed by the union of two large vitelline 

 or hepatic veins (Fig. 4, V H), which run forwards from the lateral 

 parts of the yolk mass and by the sides of the hepatic diverticulum. 

 These unite in the median line to form what will afterwards be the 

 sinus venosus, into which open the two Cuvierian veins, V D, which 

 at present are of very small size. In front of the sinus venosus the 

 heart runs forwards and slightly to the left side as the auricular 

 portion, then bends over to the right side to form the ventricle, H V, 

 and then back again towards the median plane to become the truncus 

 arteriosus, which is in connection with the anterior wall of the peri- 

 cardial cavity. 



There is as yet no separation between the auricular and ventricular 

 portions of the heart, but there is a sharp constriction between the 

 ventricle and the truncus arteriosus. When the truncus arteriosus 

 reaches the anterior Avail of the pericardial cavity, it divides at once 

 into right and left branches, which run directly outwards at right 

 angles to the axis of the body, and just below the floor of the 

 pharynx. Each division of the truncus is flattened clorso-ventrally 

 and ends blindly, its extremity usually showing a more or less 

 marked tendency to divide into two or three lobes. 



4. The Blood Vessels of the Visceral Arches (Fig. 4). 



a. The Maudiluhhr Arch. — We have found no vessels at this stage 

 that can be considered as belonging to the mandibular arch. The 

 arch itself is hardly recognisable as such. 



b. The Hyoid Arch. — Along almost the entire length of the 

 hyoid arch, and occupying almost exactly its axis, there extends 

 a narrow irregular lacunar space (Fig. 4, E H'). Its dorsal end 



