122 Edward Phelps Allis jr.. 



trigemino-facialis chamber. The internal carotid arises from the 

 etferent glossopharyngeal artery slightly dorsal to the external carotid, 

 and it is said to thicken, a short distance from its origin, into a 

 gland-like structure which represents the remains of the pseudo- 

 branch. The internal carotid is said to pass directly through this 

 "rete mirabile'", from end to end^ and. although nothing is said about 

 it, the artery must certainly give one or more branches to the 

 structure, for according to the descriptions the structure has no other 

 possible aiferent arterial connection. From the structure the 

 ophthalmica magna and three small encephalic arteries arise, aside 

 from the anterior, orbito-nasal prolongation of the internal carotid 

 itself. The anterior and posterior encephalic arteries each anastomose, 

 inside the cranial cavity, with their fellow of the opposite side, but 

 whether or not one or the other of these two anastomoses is the 

 homologue of the intramyodomic anastomosis of the internal carotids 

 in the Loricati can not be told. McKenzie apparently thinks not, 

 for he says there is no circulus cephalicus in the üsh. Eidewood 

 (1899), however, shows the internal carotids of Silurus apparently 

 regularly united by commissure, as in other teleosts. 



The conditions in Ameiurus would accordingly seem to indicate a 

 line of descent for these vessels other than that that leads through 

 Amia to teleosts. For, when the arteria hyoidea, or mandibular 

 aiferent artery, had for some reason or other ceased to give an 

 adequate blood supply to the pseudobranch, that organ apparently 

 acquired the necessary supply, in Ameiurus, by way of its primary 

 connection with the internal carotid, instead of, as in xbnia, by way 

 of a secondary connection with the external carotid. And, in 

 acquiring this relation to the internal carotid, the pseudobranch 

 has apparently first become sessile on that artery, and has then 

 enveloped it, so that the ophthalmica magna and encephalic arteries 

 a,ppear to arise directly from the pseudobranch instead of from the 

 efferent mandibular and internal carotid arteries respectively^ The 

 vessels in Gadus may perhaps, as already stated, be related to this 

 evident arrangement in Ameiurus, rather than to that in the Loricati ; 

 as are probablj' also those in Esox lucius (Mauker, 1884). 



In Fig. 12 the arrangement of the vessels, as given by F. W. 

 Müller (1897), in a 26 mm larva of Lcpidosieus is shown. Here 

 the afferent hyoidean and mandibular arteries are both retained, 

 and there is both an opercular gill and a mandibular one. A com- 

 missure connects the hyoidean and mandibular aortic arches. The 



