An almost Perfect Skeleton of Pareiasaurus serridens. 129 



wards and outwards and slightly backwards. The prezygapophysis 

 is united to a large extent with the transverse process. 



On passing backwards the height of the vertebras is slightly 

 increased, the 18th vertebra measuring •220m., and the body 

 •080m., while the large articulation for the rib becomes less 

 developed, and at the 18th vertebra has an articular surface 

 measuring only •080iti. The vertebra, however, is larger across 

 the postzygapophyses (•215m.), but owing to the smaller size of 

 the ribs it only measures •lOOm. across the transverse processes. 



In the 19th and 20th vertebrae the articulations for the ribs are 

 even smaller than in the 18th, but the vertebrge otherwise closely 

 resemble the 18th. 



Intercentra have apparently been present between all the dorsal 

 vertebrae. 



The sacrum consists of two vertebrae, the 21st and 22nd of the 

 series. Seeley in his first paper on Pareiasaurus says of the sacrum 

 in P. homhidcns (p. 84) : " Two vertebrge are anchylosed together in 

 the sacral region, but only the first contributes to the support of the 

 pelvis." In his second paper, while he states that there are 19 pre- 

 sacral vertebrae in his second specimen of P. bombidens, he makes no 

 reference to the sacrum, but in P. baini he found no fewer than four 

 sacral vertebrge. These four, however, he divides as follows : the 

 first a sacro-lumbar, the second a true sacral, and the last two sacro- 

 caudal. This explanation is, probably at least, in part correct. In 

 Pareiasaurus serridens the two sacral vertebree are anchylosed 

 together, and each gives support to the ilium. The pleura- 

 pophysis of the first sacral is of enormous size and somewhat 

 similar to that figured in P. bombidens. From its articulation with 

 the body of the vertebra to its lowest point it measures •200m. It 

 gives a very large articulation to the ilium, measuring •225m. in 

 length and -OSOm. in breadth. The second sacral rib is com- 

 paratively small and articulates with the ilium behind the upper 

 part of the 1st sacral rib. It is possible that the 20th vertebra, the 

 last of the dorso-lumbar series, had a rib which lay inside of the 

 iliac crest and may even have been attached to the ilium by liga- 

 ment, but as the body has been freely movable on the body of the 

 sacral vertebra, this vertebra cannot in P. serridens be regarded a& 

 anything but a dorso-lumbar. The 23rd vertebra has also been in 

 no way fixed to the sacrum, so that, even though it may have had a 

 rib which approached the ilium, it cannot be regarded as a sacral 

 vertebra. It may therefore be taken as established that Pareiasaurus 

 has two true sacral vertebrge, but that in some species, e.g., P. bainiy 



