ZOOLOGY. s 245 
series with a higher tendency or terminus, thus retrograding in 
one sense. This is seen in the shortened sacrum, pleurodontu den- 
tition, etc. 
The Crocodilia of the Jurassic do not possess the ball and 
socket-jointed vertebra of the recent genera, and exhibit the plane 
articular faces of all the Jurassic and Triassic Reptilia. Their 
basicranial region is also plane like that of other orders, instead of 
vertical as in the recent forms. The Triassic Crocodiles are still 
more generalized. Their ribs are extended to the pelvis, as in ~ 
Dinosauria and Anomodontia: there are often three sacral verte- 
bre, an approach to the long sacrum of the same orders. The 
femur, with third trochanter, is an approach to that of the Dino- 
sauria; and finally the position of the nostrils near the orbits (Be- 
lodon) is a Sauropterygian feature. In the Sauropterygia the 
shortened vertebral column, and long muzzle (Pistosaurus) in the 
oldest types (Triassic), are approximations to the Crocodilia. The 
‘Dinosauria display an increasingly Crocodilian character as we 
pass into the Triassic period. The femur (Paleosaurus, Megadac- 
tylus) loses the bird-like head, and assumes the ill-defined convexity 
of the Rey Seeaclc ae the tibia (Plateosaurus) loses the bird-like 
“spine,” or crest. The ilium is shorter (Paleosaurus). Every 
student of the subject knows how much more difficult is the sepa- 
ration of the bones of Sauropterygia, Crocodilia, Anomodontia, and 
Dinosauria, of ‘the Trias, than those of the Cretaceous. There are 
types allied to the Rhynchocephalia, whose systematic position is 
doubtful, owing to the generalized character of the parts we pos- 
sess. Thus the Rhynchosaurus of the Trias of England is allied 
to that order, and to the Anomodontia. The Rhopalodon of the 
Permian has a large canine tooth, like the single one possessed by 
the Anomodontia; but with others associated, like those of the 
Rhynchocephalia. The Triassic Sauropterygia and Rhynchoce- 
phalia also agree in the anterior production of the pterygoid bones 
between the palatines to the vomer. Compare, for this point, Hy- 
perodapedon and Nothosaurus. 
We learn from such considerations as the above, and similar 
ones derived from the study of the Mammalia, that the successional 
relation of the faunz of the periods in geologic time is more strik- 
ingly exhibited by the subordinate contents of the orders than 
by the orders themselves, in relation to each other. From this we 
decide that we must look for the origin of the orders in periods 
prior to those in which we now know them, if, as some suppose, 
