630 COMPARATIVE ANATOMY 



clearly understood that the so-called ancestors are near, not on, the 

 family line. 



In such a phylogenetic series or family tree for vertebrates, morpholo- 

 gists agree that protozoa and coelenterates are near the root. In addition 

 nearly all are convinced that in many respects flatworms lie near the 

 main line of ancestry. Great diversity of opinion, however, prevails in 

 regard to the groups which may be thought to link flatworms to verte- 

 brates. There is great uncertainty, even, as to the group which may be 

 considered as immediately ancestral to vertebrates. 



Among the immediate precursors of vertebrates have been included 

 arthropods, annelids, urochordates, hemichordates, cephalochordates, 

 and even the trochophore larva of annelids. The evidence which has 

 been advanced by the proponents of these various theories may be briefly 

 summarized. 



ANNELID THEORY 



Since vertebrates are strikingly metameric, it is not surprising that 

 early speculations generally assumed that vertebrates have descended 

 from metameric invertebrate ancestors. Saint Hilaire (1822) sought to 

 derive vertebrates from insects. The necessity of reversing the upper and 

 lower sides of the arthropod in order to bring the nervous system on the 

 dorsal side as in vertebrates has not deterred morphologists from making 

 such comparisons. 



As phylogenetic speculation became more restrained, however, the 

 origin of vertebrates from such highly specialized forms as insects began 

 to seem less probable. Biologists, therefore, sought for more generalized 

 invertebrates as hypothetical ancestors. To some, annelids seemed to 

 meet the necessary requirements. Delsman, a recent supporter of the 

 annelid hypothesis of vertebrate ancestry writes, (1922) "Hardly two of 

 the great subdivisions or phyla of the animal kingdom show such a close 

 agreement, even in the details of their structure, as annelids and verte- 

 brates, and in no other case can the structure of one be derived so com- 

 pletely from that of the other." 



In 1875, Semper and Dohrn attempted independently to demonstrate 

 the origin of chordates from annelids. 



Both annelids and chordates are coelomate and metameric; both 

 elongate by means of terminal growth; circulatory and urogenital systems 

 are fundamentally similar; resemblances of nervous and muscular systems 

 may be found; and the differences between annelids and chordates may 

 be interpreted as coenogenetic. 



The reversal of upper and lower sides necessary to convert an annelid 

 into a vertebrate is not regarded by supporters of the theory as a serious 

 difficulty. They point to the fact that some animals swim either side 



