MERBIAM: TK1AHSTC Tf'ITTIIYORArKTA. 



7f> 



16. Neck short, usually with two or more anterior 



cervical centra fused. 



17. All discrete cervical and caudal intercentra 



apparently lost in some late forms, ele- 

 ments of lower arches not united ven- 

 tral ly. 



1S. Zygapophysial facets of vertebrae reduced 

 or brought into one plane and united. 



1(1. Neural spines of vertebrae strongly flat- 

 tened laterally. 



20. Posterior region of tail bent downward 



sharply for support of lower lobe of 

 fluke. 



21. Neural arches of caudal region reduced 



near bend of tail, owing to change of 

 form of fluke. 



22. Scapula narrowed distally. 



23. Episternum T-shaped instead of deltoid. 



24. Abdominal ribs often not united along 

 median line. 



16. Neck region short, but anterior cervical 



centra all separated by wide interverte- 

 bral spares. 



17. Cervical intercentra few and small. Cau- 



dal intercentra large and united ven- 

 trally to form Y-shaped lower arches. 



IS. Zygapophysial faces in the anterior half 

 of the vertebral column neither united 

 nor brought into the same plane. 



1!). Neural spines averaging relatively thick. 

 and may be circular in cross-section in 

 the anterior half of the vertebral col- 

 umn. 



20. Distal region of tail showing relatively less 



marked downward curvature. Some- 

 times bent upward strongly, and then 

 downward. 



21. Neural arches in middle third of caudal 



region not reduced. In some cases the 

 arches in this region are much elon- 

 gated and erected to broaden this 

 region of the tail. 



22. Scapula expanded distally. 



23. Episternum approaching a deltoid rather 



than T-shaped form where known 

 (Mixosattriis). 



24. Abdominal ribs heavy, and united in a 



prominent spine on the median line. 



It will be noted that in the extended series of differences which appear in 

 the table above, nearly all are features of the skeleton which have been con- 

 sidered as standing in close relationship to aquatic adaptation. These 

 include also the principal ordinal characteristics distinguishing the group. 

 The last three characters only are not considered as necessarily standing in 

 close relation to aquatic adaptation, but possibly even these are to be ascribed 

 to the effect of adaptation of this nature. In the contrast between the earlier 

 and the later ichthyosatirians it is seen that nearly all of the characters of 

 the later forms which are considered as standing in close relationship to aquatic 

 adaptation are less distinctly expressed in the Triassic representatives of the 

 group. In other words the characteristics of the earlier forms are practically 

 all nearer those of shore or land reptiles, and those of the later ones take the 

 direction of specialization toward an adapted fish-like form. The evidence 

 before us shows that the trend of the specialization of the Ichthyosauria has 



