336 



continues dorsal to all the clefts as far as my sections went. A branch 

 is sent downward from this dorsal artery, and upward from the ventral 

 artery, immediately anterior to each cleft ; the branches unquestionably 

 connecting with each other, either directly or through capillaries, though 

 this could not be determined in my sections. 



My next stage is marked as a 40 mm embryo, and of it I have 

 a very perfect series of sections. The vertebralis impar capitis artery 

 here lies dorsal to the pharynx, in the region of the velar folds; be- 

 ginning at the anterior end of the velar region, where the folds arise 

 from the lateral surface of the pharynx, and extending backward 

 almost to the hind end of the velar region, where the folds are sus- 

 pended by a narrow median dorsal pedicle. This artery passes be- 

 tween j and hence dorso-mesial to the anterior connecting velar 

 processes of Ayers and Jackson's descriptions, and, immediately 

 anterior to the anterior median process of the suprapharyngeal plate, 

 separates into two parts. Shortly before the impar artery thus separates 

 into these two parts, a median branch is sent downward into the 

 pedicle of the velar folds. The two main arteries then run backward 

 and laterally, ventral to the suprapharyngeal plate, one on each side 

 of the anterior median process of that plate. They soon each send 

 one branch forward and another backward along the ventral surface 

 of the spinal cord, the two branches forming a continuous line and 

 seeming almost as parts of a single artery that here would intersect 

 the main artery. In my one embryo one of the two branches that are 

 here sent backward gradually disappeared in the sections, the other 

 gradually increasing in size and gradually assuming a median position 

 beneath the cord. Vertebral branches are sent from it, on either side, 

 to the dorsal trunk muscles. 



The main artery, on each side, now continues backward along 

 the lateral surface of the pharynx, lying always dorsal to the branchial 

 clefts. Posterior to the seventh cleft, the artery, on either side, sends 

 a large branch, if it be not the main artery itself, rnesially, to unite 

 in the median line with the median dorsal aorta. Beyond this point the 

 two arteries were not traced. These two large lateral arteries, forward 

 to the point where they each give off an anterior branch, are the 

 common carotids of Müller's descriptions of Bdellostoma heterotrema. 

 The anterior branch is the external carotid of the same descriptions ; 

 the main artery beyond that point, and up to the point where it joins, 

 in the middle line, its fellow of the opposite side, being the internal 

 carotid. This identification of the several carotid arteries seems to 

 me wrong ; the external and internal carotids being probably represented 



