Vol. 69.] SKELETON OP ORKITHODESM-US J.ATll)I-:_X.s. 413 



diversified in their nature, that the diagnosis of constant characters 

 is rendered extremely difficult. The fusion of the dorsal vertebrae 

 into a notarium in Ornithostoma and Ornithodesmus is absent in 

 BhamphorhyncTius, Pterodactylus, and all other known types. The 

 anchylosed shield-shaped sacrum is permanent throughout the 

 sub-order Rhamphorhynchoidea, Ornithostoma and Nyctosaurus 

 being of this type. In many of the examples of Pterodactylus the 

 form of the sacrum is not definitely determinable. If all follow 

 the type of Pt. suevicus, they possess a sacrum approaching the 

 fused shield-like sacrum of the Rharnphorhynchoidea. Whether 

 the sacral ribs in Scaphognathus, Dimorphodon, and Orniihodesnnis 

 are free or blended is yet obscure. The number of vertebrae com- 

 prising the sacrum is variable. In the llhamphorhynchoidea the 

 sacrum is avian, shield-shaped, with anchylosed ribs ; and in the 

 Pterodactyloidea it is avian or reptilian, with the ribs either free 

 or anchylosed. 



As fresh discoveries arise, we find that the division into long- and 

 short-tailed groups is not a good one. Scaphognathus crassirosiris 

 has been placed in the long-tailed group by authors, as the wing- 

 metacarpal is half the length of the antebrachium, and therefore 

 Sc.purdoni has followed ; but in neither is the tail known. Goldtuss 

 restored the former with a short tail, and Zittel thought this correct. 

 To the same sub-order Scaphognathoidea belong Scaphognathus 

 with a short tail and Dimorphodon with a long; and to the Rham- 

 phorhynchoidea, Ornithostoma and Nyctosaurus with a short tail, 

 and lihamphorliynchus with a long. In the Pterodactyloidea only 

 short-tailed forms occur. 



The form of the sternum in Scaphognathus crassirostris is un- 

 certain, on account of the position and state of its preservation; 

 and this element in Sc. purdoni and Dimorphodon is undiscovered. 

 The sternum of Ornithodesmus is too highly specialized to provide 

 any safe guide to the probable form of that bone in the two first- 

 named genera. 



The type of sternum of lihamphorliynchus, with its strong 

 anterior styliform process possessing no true keel, and the sternal 

 plate a broadly-expanded shield with square anterior borders, is 

 not only common to this long-tailed genus, but also to the short- 

 tailed Ornithostoma and Nyctosaurus, therefore invalidating this 

 type of sternum. as a peculiar feature of the long-tailed group. In 

 Pterodactylus, as exemplified by Pi. suevicus, there is an anterior 

 spine, but no true keel ; and the sternal plate i3 semi-elliptical, 

 with its anterior borders rounded. 



The scapula and coracoid in the early genera of the Scapho- 

 gnathoidea were not fused ; but later they became so, as shown by 

 Ornithodesmus, and in the Rhamphorhynchoidea by Ornithostoma. 

 In the Pterodactyloidea no genus is known in which these bones 

 are anchylosed. 



The humerus is crushed in all the type-specimens, and this may 

 have had some effect in splaying out the deltoid crest: for probably 

 in all it had originally a slight inward curve. Such a curve is seen 



