THE DIGESTIVE SYSTEM 297 



Beard and Von Kupffer are persuaded that vertebrates have had two 

 mouths — an old paleostoma, and new neostoma. The paleostoma, in 

 their opinion, may be represented by the hypophysis, which in some 

 Cyclostomes, e.g. Bdellostoma, opens directly into the pharynx. Further- 

 more, the presence of a pre-oral gut in vertebrate embryos is interpreted 

 as a support of this hypothesis. Von Kupffer thinks that the hypophysial 

 mouth may be homologous with the mouth of Urochords, while Dohrn 

 assumes that it is comparable with the mouth of annehds. Such views 

 have seemed to morphologists to have a rather insecure support, and have 

 therefore been regarded as speculations. (Figs. 248, 250) 



That the chordate mouth is not homologous with the mouth of inverte- 

 brates has generally been assumed, especially by those who have upheld 

 either the annelid or the arachnid hypothesis of vertebrate ancestry. 

 Each of these theories seems to need to assume a reversal of dorsal and 

 ventral surfaces. But were such a reversal to occur, the ventrally situated 

 mouth of the invertebrate would take a dorsal position, and would con- 

 sequently He dorsal to the central nerve cord, whereas in all chordates 

 the mouth is ventral to the central nerve cord. Hence the assumption 

 is made that a new ventral mouth must have arisen when the worm turned 

 over and became a chordate. Minot has sought to avoid this theoretical 

 necessity by assuming that the lateral halves of the supra-esophageal 

 ganghon of the annelid became separated and converted into the eyes 

 of vertebrates, while the old mouth migrated to its present position on 

 the ventral side of the vertebrate, which is the former dorsal side of the 

 invertebrate. 



Whether we accept or reject either the annehd or the arachnid theory 

 of the origin of vertebrates, we must believe that there have been at least 

 two mouths in the course of vertebrate phylogenesis. The reason for this 

 conclusion is that the original coelenterate mouth becomes the mouth in 

 no vertebrate, while only in Cyclostomes, dipnoi, and possibly some 

 amphibia does it become the anus. But the coelenterate mouth becomes 

 the blastopore of chordate embryos. And the blastopore of chordates 

 lies at the posterior end of the body and forms the neurenteric canal, 

 which connects the neural tube with the enteron, while the chordate 

 mouth develops at the anterior end of the enteron. Consequently, it 

 seems indisputable that there have been at least two mouths in the history 

 of vertebrates. 



While, however, all agree that the vertebrate mouth is not the primary 

 animal mouth, and that at least two mouths have successively appeared, 

 some morphologists believe that there have been at least three mouths. 

 Delsman (1922) reviving an earlier suggestion of Kowalevsky (1877) 

 claims that "in the ontogeny of vertebrates we see three successive 

 mouths appear in the same succession as they appeared in phylogeny, 



