92 JOHAN KIÆR. M.-N. Kl. 



Stage, this agreement might also be explained as proof of descent from a 

 common ancestor. 



Looking back upon the characters that have been considered we find 

 that the following by their divergence argue most against the direct descent 

 of Paradoxidae from Mcsonacidac: 



1. The course of the facial sutures, but in particular their obliteration 

 (symphysis) in Mesonacidae. 



2. The structure of the glabella, especially the development of the furrows. 



3. The evolution of the palpebral lobes. 



4. The pleural furrows of the thoracic segments. 



5. The sculpture of the shell. 



6. The nepionic stages in the ontogenetic development, with the excep- 

 tion of the intergenal spines. 



The following characters entirely agree, and may be taken as indica- 

 tions of direct descent: 



1. The structure of the hypostome, and its attachment to the doublure of 

 the cranidium. 



2. The protaspis stage and the intergenal spines in the nepionic stages^ 



In addition, there is the usual form of the body, which in its main 

 features entirely agrees in both of the above. These points of resemblance,, 

 however, may also be ascribed to common ancestry and convergent evolution. 



As we see, most of these factors are in disfavour of the hypothesis 

 that Mesonacidae continued in the Middle Cambrian Paradoxidae. 



From our present knowledge, they must be regarded as 

 two parallel families that have developed i ndependentl^^ of 

 each other from common ancestors in the oldest part of the 

 Lower Cambrian or at a still earlier period. 



This view of their relationship agrees very closely with the circum- 

 stance that the reduction of the thoracic segments (as regards number) is 

 more advanced in the youngest Mesonacidae, than in the oldest Paradoxidae. 

 Inded we might have expected the reverse if Walcott's views were correct. 



We have also thereby an explanation as to why no one has succeeded 

 in finding undoubted transition forms between these two prominent families. 

 According to Walcott the reason is that the strata in the Middle Cam- 

 brian of the Atlantic province, in which this evolution is assumed to have 

 taken place, are incomplete. He admits, however, that the Protolenus- 

 fauna in New Brunswick and in Shropshire are a transition fauna, although 

 they contain no trace of forms that might be regarded as missing links. 



