16& IOWA GEOLOGICAL SURVEY 



known only in the modern fauna, yet the evidence of compara- 

 tive anatomy forces us to conclude that it has come down to us 

 practically unchanged from Permian times, and that its imme- 

 diate ancestors gave rise to all reptiles with two cranial arches 

 (Archosauria or Diapsida), and possibly to a Dinosaur-avian 

 offshoot as well. Similarly, if modern Ceratodonts can be shown 

 to possess a more primitive organization than Palaeozoic Cteno- 

 dipterines and Arthrodires, with which groups they are evi- 

 dently related, it becomes a logical necessity for us to suppose 

 the more primitive group to have antedated and perhaps even 

 to have given rise to the more highly specialized. The truth of 

 this hypothesis does not require confirmation by positive evi- 

 dence such as might be furnished by the palseontological record, 

 its validity being established upon well ascertained principles 

 of comparative anatomy. Our concern is neither to impugn nor 

 to exalt the adequacy of the palseontological record. We have 

 merely to take it as we find it, and where its continuity is broken, 

 characters obliterated, and the chain of organic forms inter- 

 rupted, there is no recourse but to fill in the lacunae as best 

 we may through exercise of the trained imagination. 



What weight should be assigned to Dean's third objection 

 depends upon whether or not we adopt the view that Arthrodires 

 and Ostracophores are closely related, and that both are dis- 

 tinct from Pisces proper. The problem may be still unsolved, 

 yet it must be remarked that very few morphologists favor a 

 separation of Arthrodires from ordinary fishes, and the idea 

 that "bothriolepids and cephalaspids " share close affinities with 

 Coccosteus-like forms may be likened to a goal that is unat- 

 tainable except after having penetrated a Daedalian labyrinth 

 of uncertainties and possibilities. It may be pertinent- to recall, 

 furthermore, that Patten's recent studies of Bothriolepis have 

 convinced him that the lowly group to which it belongs should 

 be separated further than ever from true fishes, and elevated to 

 the rank of a new and independent class. This implies, of 

 course, an effectual separation between the two groups which 

 Dean and Hussakof unite in their understanding of the term 

 Placodermata. We may close this phase of the discussion by 

 quoting Professor Patten's latest utterance in regard to the 



