PEDIGREES AND RELATIONSHIPS 



-j 



ities, ought to be split into two separate orders, namely the 

 Dinosauria proper, as typified by the carnivorous and thecodont 

 Megalosaurus and its allies, and the Orthopoda, to include the 

 herbivorous types, in which the teeth have more or less com- 

 pletely lost the original socketed character. Whichever view 

 be adopted, it is evident that all these reptiles are very closely 

 related, and that it is the carnivorous group which connects 

 them with the belodonts (Parasuchia). With that group 

 dinosaurs agree in their double temporal arcades, fixed quadrates, 

 and the inclusion of the pubis in the acetabular cavity of the 

 pelvis, although they have lost the clavicles. The more specia- 

 lised dinosaurs, like crocodiles, have developed ball-and-socket 

 articulations to the vertebrae, but it is remarkable that the posi- 

 tions of the balls and sockets are just the reverse of those in 

 the latter group, the socket being at the hind and the ball at 

 the front end of the vertebral column, thus forming the opistho- 

 ccelous (in contradistinction to the proccelous) type of verte- 

 bral structure. 



A few words may be devoted in this place to the supposed 

 direct phylogenetic relationship between birds and dinosaurs. 

 When the many curious resemblances undoubtedly existing 

 between the bones of the pelvis and hind-limbs of those dino- 

 saurs which habitually assume the upright posture, such as the 

 incipient union between the lower end of the tibia and the as- 

 tragalus, and the consequent tendency towards the formation of 

 a tibio-tarsus and a tarso-metatarsus, were first brought to 

 light, it was confidently assumed that dinosaurs represented 

 the ancestral stock from which birds had originated. This 

 view was strengthened by the fact that primitive birds were 

 furnished with teeth implanted in sockets after the megalo- 

 saurian type. Neither was it weakened by the circumstance 

 that birds have a free quadrate-bone and only a single (lower) 

 temporal arcade ; the loosening of the attachment of the 

 quadrate being doubtless correlated with the loss of the upper 

 temporal arcade. A more serious, although not insuperable 

 objection, is the retention of clavicles by birds. 



When, however, the doctrine of parallelism in development, 

 owing to structural adaptations for more or less nearly similar 

 modes of life, became generally recognised as a powerful factor 

 in the evolution of animals, opinion veered in the opposite 



