86 MKMOIRS OF THE UNIVERSITY OF CALIFORNIA. 



but in which an original bieipital rib articulation had persisted. The European 

 M.( ?) atavus may be in part related to Mi.rowiu>-HN, and in part to other groups 

 with bieipital rib articulation. 



It is ckmbtful whether any single group of characters should be used in 

 attempting to determine the course of development or the relationships of the 

 forms in this order. We may, however, fix upon those features which have- 

 been most persistent, and have generally followed the most direct course of 

 evolution. The most persistently differing characters in close relation to 

 adaptation are found in the variation in structure, and possibly in function of 

 the paddles. This variation is apparent from the Triassic forms up to the 

 acme of the Jurassic representatives of the group. The most marked differ- 

 ence in characters Avhich might be considered as fundamental is found in the 

 rib articulation. Both the broad and narrow types of paddles are far removed 

 from a primitive form of the limb. While the rib articulation has evidently 

 also varied considerably in all of the forms, it is to be presumed that in certain 

 parts of the vertebral column the characters have changed less than in the limits, 

 which are in closer touch with the environment. 



If an attempt be made to classify all of the ichthyosaurian groups on the 

 basis of the structure of the paddles or of the rib articulation, it is seen that 

 the characters do not vary together. They are so entangled that broad and 

 narrow types of paddles appear in forms with mainly bieipital articulation, 

 as in the latipinnate and longipinnate forms of Ichthyosaurus, and are also 

 found in types with mainly single articulation as in the broad-paddled Mixo- 

 saurus and the narrow-paddled Merriamia. If therefore any phylogenic 

 scheme is to be based on these characters, it will be necessary to determine 

 which of these two kinds of variation is the more primitive, and which has 

 developed secondarily in the separate lines of descent. Of the two sets of 

 characters, the nature of the rib articulation would seem to be more nearly 

 fundamental than the structure of the paddles; and one of the forms of rib 

 articulation is presumably primitive, the other being a variation from it. 



From evidence which has been presented in the general discussion of the 

 vertebrae (p. 53) there seem to be some reasons for considering the double 

 articulation as more probably primitive than the single articulation. If such 

 be the case, all Triassic forms excepting Toretocnemus and M.(l^) iititrtix 

 must be set aside as specialized and not ancestral to Ichthyosaurus. If the 

 double articulation be primitive, the forms with bieipital articulation, as 

 Ichthyosaurus and Baptanodon, have retained a primitive character not pre- 

 served in the shastasaurines or in Mixosaurus, and these groups must be con- 

 sidered as specialized branches from the persistent stock, and not ancestral to 

 Ichthyosaurus. 



On the other hand, if the single articulation of the rib head be primitive, 



