86 ■ PKOP. H. G. SEELEY OIT 



As early as December 1874* I suggested the name Cetiosauria 

 as an ordinal division of the Dinosauria, and therefore am in har- 

 mony with Professor Marsh in separating these types from the re- 

 mainder of the Dinosauria, though the separation may be based 

 upon different data from those adopted by Professor Marsh. These 

 animals are characterized by pneumatic vertebrae. And therefore 

 when an additional order is instituted for animals with cavernous or 

 pneumatic vertebraB, the Theropoda of Marsh t, under which Coelurus 

 is grouped, it becomes necessary, in order to determine the system- 

 atic position of Thecospondylus, to briefly review its relations to 

 allied animals. I have no doubt that the two ordinal groups Sauro- 

 poda and Theropoda should be united into one order, the Saurischia, 

 while the Stegosauria and Ornithopoda should be united into an order, 

 the Ornithischia. 



If Prof. Marsh is correct in attributing to Coelurus the pelvic bones 

 which have been figured (Amer. Journ. Science, vol. xxvii. pi. xi., 

 April 1884), then there can be no doubt that the genus correctly 

 finds its place in the Saurischia, in some such position as its dis- 

 coverer suggests ; for its pubis does not show a generic difference of 

 form from Allosaurus. But in his account of the vertebral column 

 (Amer. Journ. Science, vol. xxi., April 1881) it is said that the ribs 

 preserved have undivided heads, and in the figure of a dorsal ver- 

 tebra (Z. c. pi. xi. fig. 2 a) there is no sign of a capitular facet on 

 the neural arch, while the fractured, compressed, transverse process 

 is subcrocodilian. This is a remarkable divergence from the divided 

 rib- articulation hitherto found in the Ornithischia; but is an in- 

 telligible modification of the Saurischian type. Its chief interest 

 lies in the rib being attached to the transverse process as in the 

 Oruithosaurian genus Pachyrhamphus^ with a vertical compression 

 of the process, which is Crocodilian or Avian. But while the 

 Saurischia may thus be exceptionally modified in vertebral cha- 

 racters, the pneumatic skeleton is an approximation towards Orni- 

 thosaurs and Birds, which may be a stronger evidence of organic 

 affinity. And it serves to demonstrate the way in which the 

 pneumatic skeleton was developed, in animals which not only did not 

 fly, but show no sign of degeneration from flying types. It 

 cannot therefore be classed as an adaptive modification of structure 

 consequent on flight. Hence it may be regarded as indicative of a 

 community of plan of vital organs so far as the respiratory system 

 is concerned. It is curious that while the Saurischia thus make 

 an approximation to Ornithosaurs, and the pelvic structures and 

 hind limb to some extent support the comparison, — the pelvic 

 structures and the hind limb, in the Ornithischia, on the other 

 hand, rather approximate towards Birds in plan of the skeleton, 

 although the pneumatic condition of the vertebrae is not a character 

 in common between them. 



* Quart. Journ. Geol. Soc. vol. xxx. p. 690. 



t Amer. Journ. Sci. vol. xxi. p. 423 (May 1881), vol. xxiii. p. 84 (Jan. 1882), 

 vol. xxTii. p. 337 (April 1884). 



