1867.] PROF, HUXLEY ON THK CLASSIFICATION OF BIRDS. 417 



pass* are wholly, or in part, bounded by vertebrcc provided with 

 thick and expanded ribs ; and these ribs are connected, more or less 

 extensively, on the one hand, with the bodies of these vertebrse, and 

 on the other with the iliac bones. The vertebrae in question, of 

 which there are ordinarily two, constitute the sacrum. In Birds the 

 arches of the vertebrae which correspond with these in their rela- 

 tion to the nerves (and therefore must also be termed " sacral ") give 

 off comparatively slender transverse processes, which seem to answer 

 to those which unite with the tubercles of the ribs in the dorsal 

 region ; and it is by these transverse processes only that they are 

 connected with the ilia. 



4. The broad and expanded part of the sternum, which imme- 

 diately follows the coracoidal articular surfaces, receives all the 

 sternal ribs. In all recent Reptilia which possess sternal ribs, some 

 of the latter articulate with narrow prolongations, which extend 

 back from the posterior angle of the expanded rhomboidal sternal 

 plate. The sternum in Birds ossifies in a manner which has not 

 been observed in any Reptile. 



6. The ischia never unite in a median ventral symphysis ; and 

 both pubes and ischia are directed backwards, approximately parallel 

 with one another and with the spinal column. 



7. The proximal constituent of the tarsus is anchylosed with the 

 tibia into one tibio-tarsal bonef ; the distal element of the tarsus 

 similarly unites with the second, third, and fourth metatarsal bones, 

 and gives rise to the tarso-metatarsal bone. The metatarsal of the 

 hallux is shorter than the others, and does not reach the tarsus. 



Unless, as Gegenbaur has rendered probable, the hind limb of the 

 extinct reptile Compsognathus was similarly modified, these cha- 

 racters are diagnostic of birds. In any case they are highly charac- 

 teristic of them. 



8. Birds have hot blood, a muscular valve in the right ventricle, 

 a single aortic arch, and remarkably modified respiratory organs ; 

 but it is, to say the least, highly probable that the Pterosauria, if 

 not the Dinosauria, shared some of these characters with them. The 

 amount of work involved in sustaining a Pterodactyle in the air would 

 seem, physiologically, to necessitate proportional oxidation and evo- 

 lution of waste products in the form of carbonic acid. If so, a pro- 

 portional quantity of heat must have been evolved, and there must 

 have been a ready means of eliminating the carbonic acid from the 

 blood. We know of no such means, except those which are afforded 

 by highly developed circulatory and respiratory organs ; and there- 

 fore it is highly probable that the Pterodactyles had more perfect 



* The sciatic nerve of the Crocodile is formed, for the most part, by a root 

 which leaves the spinal canal by the intervertebral foramen, interposed between 

 the two sacral vertebra;, and which passes between the two expanded sacral ribs. 

 It receives a large accessory branch from the preceding, and a smaller from the suc- 

 ceeding, spinal nerve. In Gecko verus the root of the sciatic nerve, which passes 

 out between the two sacral vertebrcTc, is smaller than that which lies in front of 

 it, between the anterior sacral and the last lumbar vertebra-. 



t See Gegenbaur, /. c. 



Proc. Zool. Soc— 1867, No. XXVII. 



