4 The Value of Paleontology. 
the neck vertebra are similar to those of the former, while por- 
tions of the skull resemble corresponding ones of the latter. The 
foot of a dinosaur is intermediate between that of a reptile and 
that ofa bird; so are the sacrum and pelvis. The sternum of a 
frog of the family Déscoglosside is intermediate between those of 
ordinary frogs and salamanders ; so are the vertebre and ribs. 
Examples of the limitation of the latter rule are still more nu- 
merous. They may be produced from the three cases cited. 
Thus in the Dinosaur it might once have been said that the jaws 
did not partake of the intermediacy, because they all present 
teeth, and are never smooth, like those of birds. Yet birds with 
teeth have recently been discovered, which deprives us of the use 
of this character as a definition. In the Discoglossid frog the 
cranium is not intermediate in structure between the frog and 
salamander, but is that of a frog. In the Zoxolophodon the tooth- 
less front of the upper jaw is not a general character of either of 
the orders which it stands between. 
These difficulties arise from the existence of the subordinate 
variations or sub-types of a general or major pattern, and for their 
resolution require only a new application of the first law of uni- 
formity on the lower plane. If the sub-characters defining the 
sub-pattern be known, the existence of one presupposes that of 
the others. The structure of an artiodactyle astragalus will not 
enable me to infer the character of the incisor teeth of the 
animal; for this I require some other, more minutely correlated 
portion. SolIcan infer the ribs and vertebra from the sternum 
of the Discoglossid frog, but not the cranium; for this I require 
some part correlated with Discoglossid characters only, and not 
only significant of the relations to the orders of Batrachians, as 
are the characters mentioned, although it happens by the accident 
of discovery that none but such frogs possess them to-day. 
The two laws which further aid the deductions of the paleonto- 
logist are those of mechanical relations and of embryonic paral- 
lelism. One structure requires another in order that an animal 
be viable. Thus long legs in a grazer presuppose a long neck to 
enable it to reach the ground with its lips. Hooked claws pre- 
suppose carnassial teeth or ahooked beak. To be properly poised 
on two legs instead of four, the weight of the viscera must be 
transferred backwards and the anterior regions of the body 
