-MKRRIA.M: TRIASSIC ICHTIIYOSAURIA. 51 



to the anterior caudal region seemed actually to show a beginning division of 

 the articular face. In one case, where the evidence seemed strongest, the 

 apophysis on the vertebral centrum showed also some indication of division. 

 It should, however, be noted that even in this instance the evidence hardly 

 warrants a definite statement that double articulation occurs, as the side of the 

 vertebral centrum is badly broken up, and does not present a satisfactory 

 sin-face on which to make a thoroughly trustworthy observation. The heads 

 of these ribs arc moreover not fully exposed, and it is not impossible that fur- 

 ther preparation with a needle will show that the apparent gap between the 

 head and tubercle is filled by a bridge of bone covered by matrix lying in the 

 extension of a lateral groove. It is also worth noting that ribs with heads 

 formed as these appear to be as they are shown in the matrix, could not come 

 into direct articulation with the diapophysis without elevating the shaft above 

 the level of the other ribs. 



In the case of the American Triassic genera, careful search has been made 

 to determine whether any division of rib heads in the posterior dorsal or caudal 

 region could be discovered in any type excepting Toretocnemus. In the poste- 

 rior dorsal region the rib heads are found to be generally very broad, but show 

 no indication of incipient division. In the caudal ribs no evidence of division 

 of the head has been observed. 



From the evidence available, some doubt must exist as to the true character 

 of the rib articulation in the ancestral ichthyosaur. In all Triassic types 

 excepting Toretocnemus and certain forms of the M.(1) atavus group the 

 articulation is predominatingly unicipital. In all post-Triassic forms the artic- 

 ulation is mainly bicipital. If the double-headed articulation is primitive, 

 such forms as Cymbospondylus, Mixosaiims, Delphi i/oxtiurus, and Shastasaurus 

 are presumably not in the direct line of descent, and the later ichthyosaurs are 

 probably to be derived from some such form as ..(?) atavus or Toretocne- 

 iii UK. If the unicipital articulation is primitive either Cymbospondylus or 

 Mixosaurus might stand near the line of descent. 



As far as the known relationships of the Ichthyosauria are concerned, there 

 seem to be no particularly strong reasons for considering that either of the 

 known types of articulation must be the primitive one. Single-headed ribs 

 are perhaps the rule in the primitive Diaptosauria, but incipient bicipital ar- 

 ticulation seems to be present in Kadaliosaurus. In the Phytosauria, which 

 have been considered as probably very closely related to the Ichthyosauria, 

 the rib articulation is markedly double-headed in the cervical and dorsal re- 

 gions. If the ancestry of the ichthyosaurian group be pushed back far 

 enough so that its origin may be considered to represent as early a date as 

 that of the Diaptosauria, the ichthyosaurs must be supposed to have originated 

 from some branch not far from the original cotylosaurian stem. 



